Uses of Interface
org.tensorflow.Operand
Packages that use Operand
Package
Description
Defines classes to build, save, load and execute TensorFlow models.
-
Uses of Operand in org.tensorflow
Classes in org.tensorflow that implement OperandModifier and TypeClassDescriptionfinal classA symbolic handle to a tensor produced by anOperation.Methods in org.tensorflow that return OperandModifier and TypeMethodDescriptionOperand<?> Calls the function in an execution environment, adding its graph as a function if it isn't already present.Operand<?> Calls the function in an execution environment, adding its graph as a function if it isn't already present.static Operand<?> TensorFunction.validateDescription(Signature.TensorDescription description, Graph graph, String name, String prefix) Methods in org.tensorflow that return types with arguments of type OperandModifier and TypeMethodDescriptionAbstractGradientAdapter.apply(Graph graph, TFJ_Scope scope, GraphOperation operation, List<Output<?>> gradInputs) Calls the function in an execution environment, adding its graph as a function if it isn't already present.Calls the function in an execution environment, adding its graph as a function if it isn't already present.GraphOperation.inputs()Get the op's inputs, not including control inputs.Methods in org.tensorflow with parameters of type OperandModifier and TypeMethodDescriptionOperand<?> Calls the function in an execution environment, adding its graph as a function if it isn't already present.Operand<?> Calls the function in an execution environment, adding its graph as a function if it isn't already present.Usetinstead of the Tensor referred to by executing the operation referred to byoperand.MakesSession.Runner.run()return the Tensor referred to by the output ofoperand.Register a tensor as an input of the function.Register a tensor as an output of the function.Method parameters in org.tensorflow with type arguments of type OperandModifier and TypeMethodDescriptionCalls the function in an execution environment, adding its graph as a function if it isn't already present.Calls the function in an execution environment, adding its graph as a function if it isn't already present.Graph.completeSubgraph(Set<Operand<?>> inputs, Set<Operand<?>> outputs) Finds the operations used to produceoutputs, assuminginputsare provided.Graph.subgraphFrom(Set<Operand<?>> inputs) Get all ops that use one ofinputsdirectly or indirectly (not includinginputs), including control dependencies.Graph.subgraphTo(Set<Operand<?>> outputs) Get all ops directly or indirectly required to calculateoutputs(not includingoutputs), including control dependencies. -
Uses of Operand in org.tensorflow.framework.activations
Methods in org.tensorflow.framework.activations that return OperandModifier and TypeMethodDescriptionGets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Computes the Exponential linear unit.Exponential.exponential(Ops tf, Operand<T> input) Computes the Exponential activation function.Applies the Gaussian error linear unit (GELU) activation function with approximate set to false.Applies the Gaussian error linear unit (GELU) activation function.HardSigmoid.hardSigmoid(Ops tf, Operand<T> input) Computes the hard sigmoid activation function.Computes the linear activation function (pass-through).Applies the rectified linear unit activation function with default values.Applies the rectified linear unit activation function.Applies Scaled Exponential Linear Unit (SELU) activation functionApplies the Sigmoid activation function,sigmoid(x) = 1 / (1 + exp(-x)).Converts a vector of values to a probability distribution along the last axis.Converts a vector of values to a probability distribution.Applies the Softplus activation function,softplus(x) = log(exp(x) + 1).Applies the Softsign activation function,softsign(x) = x / (abs(x) + 1).Applies the Swish activation function,swish(x) = x * sigmoid(x).Applies the Hyperbolic tangent activation function,tanh(x) = sinh(x)/cosh(x) = ((exp(x) - exp(-x))/(exp(x) + exp(-x))).Methods in org.tensorflow.framework.activations with parameters of type OperandModifier and TypeMethodDescriptionGets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Gets the calculation operation for the activation.Computes the Exponential linear unit.Exponential.exponential(Ops tf, Operand<T> input) Computes the Exponential activation function.Applies the Gaussian error linear unit (GELU) activation function with approximate set to false.Applies the Gaussian error linear unit (GELU) activation function.HardSigmoid.hardSigmoid(Ops tf, Operand<T> input) Computes the hard sigmoid activation function.Computes the linear activation function (pass-through).Applies the rectified linear unit activation function with default values.Applies the rectified linear unit activation function.Applies Scaled Exponential Linear Unit (SELU) activation functionApplies the Sigmoid activation function,sigmoid(x) = 1 / (1 + exp(-x)).Converts a vector of values to a probability distribution along the last axis.Converts a vector of values to a probability distribution.Applies the Softplus activation function,softplus(x) = log(exp(x) + 1).Applies the Softsign activation function,softsign(x) = x / (abs(x) + 1).Applies the Swish activation function,swish(x) = x * sigmoid(x).Applies the Hyperbolic tangent activation function,tanh(x) = sinh(x)/cosh(x) = ((exp(x) - exp(-x))/(exp(x) + exp(-x))). -
Uses of Operand in org.tensorflow.framework.constraints
Methods in org.tensorflow.framework.constraints that return OperandModifier and TypeMethodDescriptionApplies the constraint against the provided weightsApplies the constraint against the provided weightsApplies the constraint against the provided weightsApplies the constraint against the provided weightsApplies the constraint against the provided weightsGets the element-wise value clipping.Calculates the norm of the weights along the axesGets the element-wise square root.Methods in org.tensorflow.framework.constraints with parameters of type OperandModifier and TypeMethodDescriptionApplies the constraint against the provided weightsApplies the constraint against the provided weightsApplies the constraint against the provided weightsApplies the constraint against the provided weightsApplies the constraint against the provided weightsGets the element-wise value clipping.Calculates the norm of the weights along the axesGets the element-wise square root. -
Uses of Operand in org.tensorflow.framework.data
Classes in org.tensorflow.framework.data that implement interfaces with type arguments of type OperandModifier and TypeClassDescriptionclassRepresents a potentially large list of independent elements (samples), and allows iteration and transformations to be performed across these elements.classRepresents the state of an iteration through a tf.data Datset.Methods in org.tensorflow.framework.data that return OperandModifier and TypeMethodDescriptionOperand<?> DatasetIterator.getIteratorResource()Operand<?> DatasetOptional.getOptionalVariant()Gets the optional variant for this DatasetOperand<?> Dataset.getVariant()Gets the variant tensor representing this dataset.DatasetOptional.hasValue()Gets the indicator of whether this optional has a value.Methods in org.tensorflow.framework.data that return types with arguments of type OperandModifier and TypeMethodDescriptionDatasetIterator.getNext()Returns a list ofOperand<?>representing the components of the next dataset element.DatasetOptional.getValue()Returns the value of the dataset element represented by this optional, if it exists.Dataset.iterator()Creates an iterator which iterates through all batches of this Dataset in an eager fashion.DatasetIterator.iterator()Method parameters in org.tensorflow.framework.data with type arguments of type OperandModifier and TypeMethodDescriptionstatic DatasetOptionalDatasetOptional.fromComponents(Ops tf, List<Operand<?>> components, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a DatasetOptional from components.static DatasetDataset.fromTensorSlices(Ops tf, List<Operand<?>> tensors, List<Class<? extends TType>> outputTypes) Creates an in-memory `Dataset` whose elements are slices of the given tensors.Returns a new Dataset which maps a function over all elements returned by this dataset.Returns a new Dataset which maps a function over all elements returned by this dataset.Dataset.mapAllComponents(Function<Operand<?>, Operand<?>> mapper) Returns a new Dataset which maps a function across all elements from this dataset, on all components of each element.Dataset.mapAllComponents(Function<Operand<?>, Operand<?>> mapper) Returns a new Dataset which maps a function across all elements from this dataset, on all components of each element.Dataset.mapOneComponent(int index, Function<Operand<?>, Operand<?>> mapper) Returns a new Dataset which maps a function across all elements from this dataset, on a single component of each element.Dataset.mapOneComponent(int index, Function<Operand<?>, Operand<?>> mapper) Returns a new Dataset which maps a function across all elements from this dataset, on a single component of each element.Constructors in org.tensorflow.framework.data with parameters of type OperandModifierConstructorDescriptionDataset(Ops tf, Operand<?> variant, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a DatasetDatasetIterator(Ops tf, Operand<?> iteratorResource, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) DatasetIterator(Ops tf, Operand<?> iteratorResource, Op initializer, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) DatasetOptional(Ops tf, Operand<?> optionalVariant, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a DatasetOptional dataset -
Uses of Operand in org.tensorflow.framework.initializers
Methods in org.tensorflow.framework.initializers that return OperandModifier and TypeMethodDescriptionGenerates the operation used to perform the initialization.Generates the operation used to perform the initialization.Generates the operation used to perform the initialization.Generates the operation used to perform the initialization.Generates the operation used to perform the initialization.Generates the operation used to perform the initialization.Generates the operation used to perform the initialization.Generates the operation used to perform the initialization.Generates the operation used to perform the initialization.Methods in org.tensorflow.framework.initializers with parameters of type OperandModifier and TypeMethodDescriptionGenerates the operation used to perform the initialization.Generates the operation used to perform the initialization.Generates the operation used to perform the initialization.Generates the operation used to perform the initialization.Generates the operation used to perform the initialization.Generates the operation used to perform the initialization.Generates the operation used to perform the initialization.Generates the operation used to perform the initialization.Generates the operation used to perform the initialization. -
Uses of Operand in org.tensorflow.framework.losses
Methods in org.tensorflow.framework.losses that return OperandModifier and TypeMethodDescriptionLosses.binaryCrossentropy(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, boolean fromLogits, float labelSmoothing) Computes the binary crossentropy loss between labels and predictions.BinaryCrossentropy.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.CategoricalCrossentropy.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.CategoricalHinge.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.CosineSimilarity.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.Hinge.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.Huber.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.KLDivergence.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.LogCosh.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.Loss.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.MeanAbsoluteError.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.MeanAbsolutePercentageError.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.MeanSquaredError.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.MeanSquaredLogarithmicError.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.Poisson.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.SparseCategoricalCrossentropy.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand the calculates the loss.SquaredHinge.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.Losses.categoricalCrossentropy(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, boolean fromLogits, float labelSmoothing, int axis) Computes the categorical crossentropy loss between labels and predictions.Losses.categoricalHinge(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions) Computes the categorical hinge loss between labels and predictions.Losses.cosineSimilarity(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, int[] axis) Computes the cosine similarity loss between labels and predictions.Computes the hinge loss between labels and predictionsComputes the Huber loss between labels and predictions.Losses.kullbackLeiblerDivergence(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions) Computes the Kullback-Leibler divergence loss between labels and predictions.Losses.l2Normalize(Ops tf, Operand<T> x, int[] axis) Normalizes along dimension axis using an L2 norm.Computes the hyperbolic cosine loss between labels and predictions.Losses.meanAbsoluteError(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions) Calculates the mean absolute error between labels and predictions.Losses.meanAbsolutePercentageError(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions) Calculates the mean absolute percentage error between labels and predictions.Losses.meanSquaredError(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions) Computes the mean squared error between labels and predictions.Losses.meanSquaredLogarithmicError(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions) Calculates the mean squared logarithmic error between labels and predictions.Computes the Poisson loss between labels and predictions.Losses.sparseCategoricalCrossentropy(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, boolean fromLogits, int axis) Computes the sparse categorical crossentropy loss between labels and predictions.Losses.squaredHinge(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions) Computes the squared hinge loss between labels and predictions.Methods in org.tensorflow.framework.losses with parameters of type OperandModifier and TypeMethodDescriptionLosses.binaryCrossentropy(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, boolean fromLogits, float labelSmoothing) Computes the binary crossentropy loss between labels and predictions.BinaryCrossentropy.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.CategoricalCrossentropy.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.CategoricalHinge.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.CosineSimilarity.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.Hinge.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.Huber.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.KLDivergence.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.LogCosh.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.Loss.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.MeanAbsoluteError.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.MeanAbsolutePercentageError.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.MeanSquaredError.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.MeanSquaredLogarithmicError.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.Poisson.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.SparseCategoricalCrossentropy.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand the calculates the loss.SquaredHinge.call(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, Operand<T> sampleWeights) Generates an Operand that calculates the loss.Losses.categoricalCrossentropy(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, boolean fromLogits, float labelSmoothing, int axis) Computes the categorical crossentropy loss between labels and predictions.Losses.categoricalHinge(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions) Computes the categorical hinge loss between labels and predictions.Losses.cosineSimilarity(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, int[] axis) Computes the cosine similarity loss between labels and predictions.Computes the hinge loss between labels and predictionsComputes the Huber loss between labels and predictions.Losses.kullbackLeiblerDivergence(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions) Computes the Kullback-Leibler divergence loss between labels and predictions.Losses.l2Normalize(Ops tf, Operand<T> x, int[] axis) Normalizes along dimension axis using an L2 norm.Computes the hyperbolic cosine loss between labels and predictions.Losses.meanAbsoluteError(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions) Calculates the mean absolute error between labels and predictions.Losses.meanAbsolutePercentageError(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions) Calculates the mean absolute percentage error between labels and predictions.Losses.meanSquaredError(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions) Computes the mean squared error between labels and predictions.Losses.meanSquaredLogarithmicError(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions) Calculates the mean squared logarithmic error between labels and predictions.Computes the Poisson loss between labels and predictions.Losses.sparseCategoricalCrossentropy(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, boolean fromLogits, int axis) Computes the sparse categorical crossentropy loss between labels and predictions.Losses.squaredHinge(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions) Computes the squared hinge loss between labels and predictions. -
Uses of Operand in org.tensorflow.framework.metrics
Methods in org.tensorflow.framework.metrics that return OperandModifier and TypeMethodDescriptionAccuracy.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Calculates how often predictions equals labels.BinaryAccuracy.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Calculates how often predictions match binary labels.BinaryCrossentropy.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the binary crossentropy loss between labels and predictions.CategoricalAccuracy.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the categorical crossentropy loss.CategoricalCrossentropy.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the crossentropy loss between the labels and predictions.CategoricalHinge.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the categorical hinge metric betweenlabelsand @{code predictions}.CosineSimilarity.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the cosine similarity loss between labels and predictions.Hinge.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the hinge loss between labels and predictions.KLDivergence.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes Kullback-Leibler divergence metric between labels and predictions.LogCoshError.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Calculates the Logarithm of the hyperbolic cosine of the prediction error.MeanAbsoluteError.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the mean absolute error loss between labels and predictions.MeanAbsolutePercentageError.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the mean absolute percentage error loss between labels and predictions.MeanSquaredError.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the mean squared error between the labels and predictions.MeanSquaredLogarithmicError.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the mean squared logarithmic error between labels and predictions.Poisson.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the Poisson loss between labels and predictions.SparseCategoricalAccuracy.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Calculates how often predictions matches integer labels.SparseCategoricalCrossentropy.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Calculates how often predictions matches integer labels.SparseTopKCategoricalAccuracy.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes how often integer targets are in the topKpredictions.SquaredHinge.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the squared hinge loss between labels and predictions.TopKCategoricalAccuracy.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes how often targets are in the topKpredictions.BaseMetric.callOnce(Ops tf, Operand<? extends TNumber> values, Operand<? extends TNumber> sampleWeights, Class<T> type) Calls update state once, followed by a call to get the resultMetric.callOnce(Ops tf, Operand<? extends TNumber> values, Operand<? extends TNumber> sampleWeights, Class<T> type) Calls update state once, followed by a call to get the resultAUC.getLabelWeights()MeanRelativeError.getNormalizer()Gets the normalizer OperandGets the current result of the metricGets the current result of the metricGets the current result of the metricGets the current result of the metricGets the current result of the metricGets the current result of the metricGets the current result of the metricGets the current result of the metricGets the current result of the metricGets the current result of the metricMetrics.sparseTopKCategoricalAccuracy(Ops tf, Operand<U> labels, Operand<T> predictions, int k) Computes how often integer targets are in the top K predictions.Metrics.topKCategoricalAccuracy(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, long k) Computes how often targets are in the top K predictions.Methods in org.tensorflow.framework.metrics with parameters of type OperandModifier and TypeMethodDescriptionAccuracy.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Calculates how often predictions equals labels.BinaryAccuracy.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Calculates how often predictions match binary labels.BinaryCrossentropy.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the binary crossentropy loss between labels and predictions.CategoricalAccuracy.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the categorical crossentropy loss.CategoricalCrossentropy.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the crossentropy loss between the labels and predictions.CategoricalHinge.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the categorical hinge metric betweenlabelsand @{code predictions}.CosineSimilarity.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the cosine similarity loss between labels and predictions.Hinge.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the hinge loss between labels and predictions.KLDivergence.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes Kullback-Leibler divergence metric between labels and predictions.LogCoshError.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Calculates the Logarithm of the hyperbolic cosine of the prediction error.MeanAbsoluteError.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the mean absolute error loss between labels and predictions.MeanAbsolutePercentageError.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the mean absolute percentage error loss between labels and predictions.MeanSquaredError.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the mean squared error between the labels and predictions.MeanSquaredLogarithmicError.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the mean squared logarithmic error between labels and predictions.Poisson.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the Poisson loss between labels and predictions.SparseCategoricalAccuracy.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Calculates how often predictions matches integer labels.SparseCategoricalCrossentropy.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Calculates how often predictions matches integer labels.SparseTopKCategoricalAccuracy.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes how often integer targets are in the topKpredictions.SquaredHinge.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes the squared hinge loss between labels and predictions.TopKCategoricalAccuracy.call(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Class<U> resultType) Computes how often targets are in the topKpredictions.BaseMetric.callOnce(Ops tf, Operand<? extends TNumber> values, Operand<? extends TNumber> sampleWeights, Class<T> type) Calls update state once, followed by a call to get the resultMetric.callOnce(Ops tf, Operand<? extends TNumber> values, Operand<? extends TNumber> sampleWeights, Class<T> type) Calls update state once, followed by a call to get the resultMetrics.sparseTopKCategoricalAccuracy(Ops tf, Operand<U> labels, Operand<T> predictions, int k) Computes how often integer targets are in the top K predictions.Metrics.topKCategoricalAccuracy(Ops tf, Operand<? extends TNumber> labels, Operand<T> predictions, long k) Computes how often targets are in the top K predictions.final OpBaseMetric.updateState(Ops tf, Operand<? extends TNumber> values, Operand<? extends TNumber> sampleWeights) Creates a NoOp Operation with control dependencies to update the metric statefinal OpBaseMetric.updateState(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Operand<? extends TNumber> sampleWeights) Creates a NoOp Operation with control dependencies to update the metric stateMetric.updateState(Ops tf, Operand<? extends TNumber> values, Operand<? extends TNumber> sampleWeights) Creates a NoOp Operation with control dependencies to update the metric stateMetric.updateState(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Operand<? extends TNumber> sampleWeights) Creates a NoOp Operation with control dependencies to update the metric stateAUC.updateStateList(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Operand<? extends TNumber> sampleWeights) Creates a List of Operations to update the metric state based on labels and predictions.BaseMetric.updateStateList(Ops tf, Operand<? extends TNumber> values, Operand<? extends TNumber> sampleWeights) Creates a List of Operations to update the metric state based on input values.BaseMetric.updateStateList(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Operand<? extends TNumber> sampleWeights) Creates a List of Operations to update the metric state based on labels and predictions.MeanIoU.updateStateList(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Operand<? extends TNumber> sampleWeights) Accumulates the confusion matrix statistics.MeanRelativeError.updateStateList(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Operand<? extends TNumber> sampleWeights) Accumulates metric statistics.MeanTensor.updateStateList(Ops tf, Operand<? extends TNumber> values, Operand<? extends TNumber> sampleWeights) Accumulates statistics for computing the element-wise mean.Metric.updateStateList(Ops tf, Operand<? extends TNumber> values, Operand<? extends TNumber> sampleWeights) Creates a List of Operations to update the metric state based on input values.Metric.updateStateList(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Operand<? extends TNumber> sampleWeights) Creates a List of Operations to update the metric state based on labels and predictions.Precision.updateStateList(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Operand<? extends TNumber> sampleWeights) Accumulates true positive and false positive statistics.Recall.updateStateList(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Operand<? extends TNumber> sampleWeights) Accumulates true positive and false negative statistics.RootMeanSquaredError.updateStateList(Ops tf, Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Operand<? extends TNumber> sampleWeights) Accumulates root mean squared error statistics.Constructors in org.tensorflow.framework.metrics with parameters of type OperandModifierConstructorDescriptionAUC(String name, int numThresholds, AUCCurve curve, AUCSummationMethod summationMethod, float[] thresholds, boolean multiLabel, Operand<T> labelWeights, long seed, Class<T> type) Creates an AUC (Area under the curve) metric.protectedCreates a MeanRelativeError metricprotectedMeanRelativeError(Operand<T> normalizer, long seed, Class<T> type) Creates a MeanRelativeError metric usingClass.getSimpleName()as the name -
Uses of Operand in org.tensorflow.framework.op
Methods in org.tensorflow.framework.op that return OperandModifier and TypeMethodDescriptionCreates an Operand that has all axes contained in the Operand's shape.MathOps.confusionMatrix(Operand<T> labels, Operand<T> predictions) Computes the confusion matrix from predictions and labels.MathOps.confusionMatrix(Operand<T> labels, Operand<T> predictions, Operand<T> weights) Computes the confusion matrix from predictions and labels.MathOps.confusionMatrix(Operand<T> labels, Operand<T> predictions, Operand<T> weights, Operand<TInt64> numClasses) Computes the confusion matrix from predictions and labels.SetOps.difference(Operand<T> a, Operand<T> b) Computes set difference of elements in last dimension ofaandbwithaMinusBset to true.SetOps.difference(Operand<T> a, Operand<T> b, boolean aMinusB) Computes set difference of elements in last dimension ofaandb.Compute the Gaussian Error Linear Unit (GELU) activation function without approximation.Compute the Gaussian Error Linear Unit (GELU) activation function.SetOps.intersection(Operand<T> a, Operand<T> b) Computes set intersection of elements in last dimension ofaandb.MathOps.l2Normalize(Operand<T> x, int[] axis) Normalizes along dimension axis using an L2 norm.Multiplies matrixaby matrixb, producinga*b.Multiplies matrixaby matrixb, producinga*b.LinalgOps.matmul(Operand<T> a, Operand<T> b, boolean transposeA, boolean transposeB, boolean adjointA, boolean adjointB, boolean aIsSparse, boolean bIsSparse) Multiplies matrixaby matrixb, producinga*b.MathOps.reduceLogSumExp(Operand<T> input, int[] axes, boolean keepDims) Computes log(sum(exp(elements across dimensions of a tensor))).NnOps.sigmoidCrossEntropyWithLogits(Operand<T> labels, Operand<T> logits) Computes sigmoid cross entropy givenlogits.NnOps.softmaxCrossEntropyWithLogits(Operand<U> labels, Operand<T> logits, int axis) Computes softmax cross entropy betweenlogitsandlabels.NnOps.sparseSoftmaxCrossEntropyWithLogits(Operand<U> labels, Operand<T> logits) Computes sparse softmax cross entropy betweenlogitsandlabels.Tensor contraction of a and b along specified axes and outer product.Tensor contraction of a and b along specified axes and outer product.Tensor contraction of a and b along specified axes and outer product.Tensor contraction of a and b along specified axes and outer product.Tensor contraction of a and b along specified axes and outer product.Computes set union of elements in last dimension ofaandb.Methods in org.tensorflow.framework.op with parameters of type OperandModifier and TypeMethodDescriptionCreates an Operand that has all axes contained in the Operand's shape.MathOps.confusionMatrix(Operand<T> labels, Operand<T> predictions) Computes the confusion matrix from predictions and labels.MathOps.confusionMatrix(Operand<T> labels, Operand<T> predictions, Operand<T> weights) Computes the confusion matrix from predictions and labels.MathOps.confusionMatrix(Operand<T> labels, Operand<T> predictions, Operand<T> weights, Operand<TInt64> numClasses) Computes the confusion matrix from predictions and labels.SetOps.difference(Operand<T> a, Operand<T> b) Computes set difference of elements in last dimension ofaandbwithaMinusBset to true.SetOps.difference(Operand<T> a, Operand<T> b, boolean aMinusB) Computes set difference of elements in last dimension ofaandb.Compute the Gaussian Error Linear Unit (GELU) activation function without approximation.Compute the Gaussian Error Linear Unit (GELU) activation function.SetOps.intersection(Operand<T> a, Operand<T> b) Computes set intersection of elements in last dimension ofaandb.MathOps.l2Normalize(Operand<T> x, int[] axis) Normalizes along dimension axis using an L2 norm.Multiplies matrixaby matrixb, producinga*b.Multiplies matrixaby matrixb, producinga*b.LinalgOps.matmul(Operand<T> a, Operand<T> b, boolean transposeA, boolean transposeB, boolean adjointA, boolean adjointB, boolean aIsSparse, boolean bIsSparse) Multiplies matrixaby matrixb, producinga*b.MathOps.reduceLogSumExp(Operand<T> input, int[] axes, boolean keepDims) Computes log(sum(exp(elements across dimensions of a tensor))).NnOps.sigmoidCrossEntropyWithLogits(Operand<T> labels, Operand<T> logits) Computes sigmoid cross entropy givenlogits.NnOps.softmaxCrossEntropyWithLogits(Operand<U> labels, Operand<T> logits, int axis) Computes softmax cross entropy betweenlogitsandlabels.NnOps.sparseSoftmaxCrossEntropyWithLogits(Operand<U> labels, Operand<T> logits) Computes sparse softmax cross entropy betweenlogitsandlabels.Tensor contraction of a and b along specified axes and outer product.Tensor contraction of a and b along specified axes and outer product.Tensor contraction of a and b along specified axes and outer product.Tensor contraction of a and b along specified axes and outer product.Tensor contraction of a and b along specified axes and outer product.Computes set union of elements in last dimension ofaandb. -
Uses of Operand in org.tensorflow.framework.op.linalg
Methods in org.tensorflow.framework.op.linalg that return OperandModifier and TypeMethodDescriptionMultiplies matrixaby matrixb, producinga*b.Multiplies matrixaby matrixb, producinga*b.MatMul.matmul(Scope scope, Operand<T> a, Operand<T> b, boolean transposeA, boolean transposeB, boolean adjointA, boolean adjointB, boolean aIsSparse, boolean bIsSparse) Multiplies matrixaby matrixb, producinga*b.Methods in org.tensorflow.framework.op.linalg with parameters of type OperandModifier and TypeMethodDescriptionMultiplies matrixaby matrixb, producinga*b.Multiplies matrixaby matrixb, producinga*b.MatMul.matmul(Scope scope, Operand<T> a, Operand<T> b, boolean transposeA, boolean transposeB, boolean adjointA, boolean adjointB, boolean aIsSparse, boolean bIsSparse) Multiplies matrixaby matrixb, producinga*b. -
Uses of Operand in org.tensorflow.framework.op.math
Methods in org.tensorflow.framework.op.math that return OperandModifier and TypeMethodDescriptionCreates an Operand that has all axes contained in the Operand's shape.ConfusionMatrix.confusionMatrix(Scope scope, Operand<T> labels, Operand<T> predictions) Computes the confusion matrix from predictions and labels.ConfusionMatrix.confusionMatrix(Scope scope, Operand<T> labels, Operand<T> predictions, Operand<T> weights) Computes the confusion matrix from predictions and labels.ConfusionMatrix.confusionMatrix(Scope scope, Operand<T> labels, Operand<T> predictions, Operand<T> weights, Operand<TInt64> numClasses) Computes the confusion matrix from predictions and labels.L2Normalize.l2Normalize(Scope scope, Operand<T> x, int[] axis) Normalizes along dimension axis using an L2 norm.ReduceLogSumExp.reduceLogSumExp(Scope scope, Operand<T> input, int[] axes, boolean keepDims) Computes log(sum(exp(elements across dimensions of a tensor))).Tensor contraction of a and b along specified axes and outer product.Tensor contraction of a and b along specified axes and outer product.Tensor contraction of a and b along specified axes and outer product.Tensor contraction of a and b along specified axes and outer product.TensorDot.tensordot(Scope scope, Operand<T> a, Operand<T> b, Operand<TInt32> aAxis, Operand<TInt32> bAxis) Tensor contraction of a and b along specified axes and outer product.Methods in org.tensorflow.framework.op.math with parameters of type OperandModifier and TypeMethodDescriptionCreates an Operand that has all axes contained in the Operand's shape.ConfusionMatrix.confusionMatrix(Scope scope, Operand<T> labels, Operand<T> predictions) Computes the confusion matrix from predictions and labels.ConfusionMatrix.confusionMatrix(Scope scope, Operand<T> labels, Operand<T> predictions, Operand<T> weights) Computes the confusion matrix from predictions and labels.ConfusionMatrix.confusionMatrix(Scope scope, Operand<T> labels, Operand<T> predictions, Operand<T> weights, Operand<TInt64> numClasses) Computes the confusion matrix from predictions and labels.L2Normalize.l2Normalize(Scope scope, Operand<T> x, int[] axis) Normalizes along dimension axis using an L2 norm.ReduceLogSumExp.reduceLogSumExp(Scope scope, Operand<T> input, int[] axes, boolean keepDims) Computes log(sum(exp(elements across dimensions of a tensor))).Tensor contraction of a and b along specified axes and outer product.Tensor contraction of a and b along specified axes and outer product.Tensor contraction of a and b along specified axes and outer product.Tensor contraction of a and b along specified axes and outer product.TensorDot.tensordot(Scope scope, Operand<T> a, Operand<T> b, Operand<TInt32> aAxis, Operand<TInt32> bAxis) Tensor contraction of a and b along specified axes and outer product. -
Uses of Operand in org.tensorflow.framework.op.nn
Methods in org.tensorflow.framework.op.nn that return OperandModifier and TypeMethodDescriptionCompute the Gaussian Error Linear Unit (GELU) activation function without approximation.Compute the Gaussian Error Linear Unit (GELU) activation function.SigmoidCrossEntropyWithLogits.sigmoidCrossEntropyWithLogits(Scope scope, Operand<T> labels, Operand<T> logits) Computes sigmoid cross entropy givenlogits.SoftmaxCrossEntropyWithLogits.softmaxCrossEntropyWithLogits(Scope scope, Operand<U> labels, Operand<T> logits, int axis) Computes softmax cross entropy betweenlogitsandlabels.SparseSoftmaxCrossEntropyWithLogits.sparseSoftmaxCrossEntropyWithLogits(Scope scope, Operand<U> labels, Operand<T> logits) Computes sparse softmax cross entropy betweenlogitsandlabels.Methods in org.tensorflow.framework.op.nn with parameters of type OperandModifier and TypeMethodDescriptionCompute the Gaussian Error Linear Unit (GELU) activation function without approximation.Compute the Gaussian Error Linear Unit (GELU) activation function.SigmoidCrossEntropyWithLogits.sigmoidCrossEntropyWithLogits(Scope scope, Operand<T> labels, Operand<T> logits) Computes sigmoid cross entropy givenlogits.SoftmaxCrossEntropyWithLogits.softmaxCrossEntropyWithLogits(Scope scope, Operand<U> labels, Operand<T> logits, int axis) Computes softmax cross entropy betweenlogitsandlabels.SparseSoftmaxCrossEntropyWithLogits.sparseSoftmaxCrossEntropyWithLogits(Scope scope, Operand<U> labels, Operand<T> logits) Computes sparse softmax cross entropy betweenlogitsandlabels. -
Uses of Operand in org.tensorflow.framework.op.sets
Methods in org.tensorflow.framework.op.sets that return OperandModifier and TypeMethodDescriptionSets.difference(Scope scope, Operand<T> a, Operand<T> b) Computes set difference of elements in last dimension ofaandbwithaMinusBset to true.Sets.difference(Scope scope, Operand<T> a, Operand<T> b, boolean aMinusB) Computes set difference of elements in last dimension ofaandb.Sets.intersection(Scope scope, Operand<T> a, Operand<T> b) Computes set intersection of elements in last dimension ofaandb.Sets.setOperation(Scope scope, Operand<T> a, Operand<T> b, Sets.Operation setOperation) Compute set operation of elements in last dimension ofaandb.Computes set union of elements in last dimension ofaandb.Methods in org.tensorflow.framework.op.sets with parameters of type OperandModifier and TypeMethodDescriptionSets.difference(Scope scope, Operand<T> a, Operand<T> b) Computes set difference of elements in last dimension ofaandbwithaMinusBset to true.Sets.difference(Scope scope, Operand<T> a, Operand<T> b, boolean aMinusB) Computes set difference of elements in last dimension ofaandb.Sets.intersection(Scope scope, Operand<T> a, Operand<T> b) Computes set intersection of elements in last dimension ofaandb.Sets.setOperation(Scope scope, Operand<T> a, Operand<T> b, Sets.Operation setOperation) Compute set operation of elements in last dimension ofaandb.Computes set union of elements in last dimension ofaandb. -
Uses of Operand in org.tensorflow.framework.optimizers
Methods in org.tensorflow.framework.optimizers with parameters of type OperandModifier and TypeMethodDescription<T extends TType>
List<Optimizer.GradAndVar<?>> Optimizer.computeGradients(Operand<?> loss) Computes the gradients based on a loss operand.Adam.createAdamMinimize(Scope scope, Operand<T> loss, float learningRate, float betaOne, float betaTwo, float epsilon, Optimizer.Options... options) Creates the Operation that minimizes the lossprotected <T extends TType>
voidOptimizer.createSlot(Output<T> variable, String slotName, Operand<T> initializer) Creates a slot in the graph for the specified variable with the specified name.Minimizes the loss by updating the variablesMinimizes the loss by updating the variables -
Uses of Operand in org.tensorflow.framework.regularizers
Methods in org.tensorflow.framework.regularizers that return OperandModifier and TypeMethodDescriptionComputes a regularization penalty from an input.Computes a regularization penalty from an input.Methods in org.tensorflow.framework.regularizers with parameters of type Operand -
Uses of Operand in org.tensorflow.framework.utils
Methods in org.tensorflow.framework.utils that return OperandModifier and TypeMethodDescriptionCasts an operand to the desired type.SparseTensor.getDenseShape()Gets the dense shape for the Sparse TensorSparseTensor.getIndices()Gets the indices for the Sparse TensorSparseTensor.getValues()Get the values for the Sparse TensorMethods in org.tensorflow.framework.utils with parameters of type OperandModifier and TypeMethodDescriptionCasts an operand to the desired type.static int[]ShapeUtils.getIntArray(Scope scope, Operand<TInt32> dims) Converts a TInt32 type Operand to a Java int arraystatic <T extends TIntegral>
long[]ShapeUtils.getLongArray(Scope scope, Operand<T> dims) Converts a TInt32 or TInt64 Operand to a java long arrayConverts a shape operand to a Shape objectConstructors in org.tensorflow.framework.utils with parameters of type Operand -
Uses of Operand in org.tensorflow.op
Methods in org.tensorflow.op that return OperandModifier and TypeMethodDescriptionCreates a 1-dimensional operand containing the dimensions of a shape followed by the last dimension.Creates a 1-dimensional operand containing the dimensions of a shape followed by the last dimension.Creates a 1-dimensional operand that represents a new shape containing the dimensions of the operand representing a shape, followed by the dimensions of an operand representing a shape to append.Ops.booleanMask(Operand<T> tensor, Operand<TBool> mask, BooleanMask.Options... options) Apply boolean mask to tensor.Ops.booleanMaskUpdate(Operand<T> tensor, Operand<TBool> mask, Operand<T> updates, BooleanMaskUpdate.Options... options) Updates a tensor at the masked values, and returns the updated tensor.Operand<?> Ops.call(ConcreteFunction function, Operand<?> argument) Calls the function in an execution environment, adding its graph as a function if it isn't already present.Flatten the shape to 1 dimension.Flatten the shape to 1 dimension.Flatten the operand to 1 dimension.Flatten the operand to 1 dimensionCreates a 1-dimensional Operand containing the Shape's first dimension.Creates a 1-dimensional Operand containing the Shape's first dimension.ShapeOps.numDimensions(Shape<TInt32> shape) Get the number of dimensions of the shape object.ShapeOps.numDimensions(Shape<U> shape, Class<U> type) Get the number of dimensions of the shape object.Creates a 1-dimensional operand containing the first dimension followed by the dimensions of the shape.Creates a 1-dimensional operand containing the first dimension followed by the dimensions of the shape.Creates a 1-dimensional operand that represents a new shape containing the dimensions of an operand representing the shape to prepend, followed by the dimensions of an operand representing a shape.ShapeOps.reduceDims(Shape<TInt32> shape, Operand<TInt32> axis) Reduces the shape to the specified axis.ShapeOps.reduceDims(Shape<U> shape, Operand<U> axis, Class<U> type) Reduces the shape to the specified axis.ShapeOps.reduceDims(Operand<T> operand, Operand<TInt32> axis) Reshapes the operand by reducing the shape to the specified axis.ShapeOps.reduceDims(Operand<T> operand, Operand<U> axis, Class<U> type) Reshapes the operand by reducing the shape to the specified axis.Get the size represented by the TensorFlow shape.Get the size of the specified dimension in the shape.Get the size represented by the TensorFlow shape.Get the size of the specified dimension in the shape.Get the size of the specified dimension for the shape of the tensor.Get the size of the specified dimension for the shape of the tensor.Removes dimensions of size 1 from the shape.Removes dimensions of size 1 from the shape.Creates a 1-dimensional Operand that contains the dimension matching the last dimension of the Shape.Creates a 1-dimensional Operand that contains the dimension matching the last dimension of * the Shape.Creates a 1-dimensional operand with the dimensions matching the first n dimensions of the shape.Creates a 1-dimensional operand containin the dimensions matching the first n dimensions of the shape.Creates a 1-dimensional operand containing the dimensions matching the last n dimensions of the shape.Creates a 1-dimensional operand containing the dimensions matching the last n dimensions of the shape.Methods in org.tensorflow.op that return types with arguments of type OperandModifier and TypeMethodDescriptionCalculate the gradients forop.Ops.call(ConcreteFunction function, Map<String, Operand<?>> arguments) Calls the function in an execution environment, adding its graph as a function if it isn't already present.RawCustomGradient.call(Ops tf, GraphOperation op, List<Output<?>> gradInputs) Calculate the gradients forop.Methods in org.tensorflow.op with parameters of type OperandModifier and TypeMethodDescriptionComputes the absolute value of a tensor.TrainOps.accumulatorApplyGradient(Operand<TString> handle, Operand<TInt64> localStep, Operand<? extends TType> gradient) Applies a gradient to a given accumulator.TrainOps.accumulatorNumAccumulated(Operand<TString> handle) Returns the number of gradients aggregated in the given accumulators.TrainOps.accumulatorSetGlobalStep(Operand<TString> handle, Operand<TInt64> newGlobalStep) Updates the accumulator with a new value for global_step.<T extends TType>
AccumulatorTakeGradient<T> TrainOps.accumulatorTakeGradient(Operand<TString> handle, Operand<TInt32> numRequired, Class<T> dtype) Extracts the average gradient in the given ConditionalAccumulator.Computes acos of x element-wise.Computes inverse hyperbolic cosine of x element-wise.Returns x + y element-wise.SparseOps.addManySparseToTensorsMap(Operand<TInt64> sparseIndices, Operand<? extends TType> sparseValues, Operand<TInt64> sparseShape, AddManySparseToTensorsMap.Options... options) Add anN-minibatchSparseTensorto aSparseTensorsMap, returnNhandles.SparseOps.addSparseToTensorsMap(Operand<TInt64> sparseIndices, Operand<? extends TType> sparseValues, Operand<TInt64> sparseShape, AddSparseToTensorsMap.Options... options) Add aSparseTensorto aSparseTensorsMapreturn its handle.<T extends TNumber>
AdjustContrast<T> ImageOps.adjustContrast(Operand<T> images, Operand<TFloat32> contrastFactor) Adjust the contrast of one or more images.Adjust the hue of one or more images.<T extends TNumber>
AdjustSaturation<T> ImageOps.adjustSaturation(Operand<T> images, Operand<TFloat32> scale) Adjust the saturation of one or more images.Computes the "logical and" of elements across dimensions of a tensor.RandomOps.allCandidateSampler(Operand<TInt64> trueClasses, Long numTrue, Long numSampled, Boolean unique, AllCandidateSampler.Options... options) Generates labels for candidate sampling with a learned unigram distribution.TpuOps.allToAll(Operand<T> input, Operand<TInt32> groupAssignment, Long concatDimension, Long splitDimension, Long splitCount) An Op to exchange data across TPU replicas.Returns the argument of a complex number.Returns the argument of a complex number.<T extends TType, U extends TType>
AnonymousMutableDenseHashTableOps.anonymousMutableDenseHashTable(Operand<T> emptyKey, Operand<T> deletedKey, Class<U> valueDtype, AnonymousMutableDenseHashTable.Options... options) Creates an empty anonymous mutable hash table that uses tensors as the backing store.RandomOps.anonymousRandomSeedGenerator(Operand<TInt64> seed, Operand<TInt64> seed2) The AnonymousRandomSeedGenerator operationRandomOps.anonymousSeedGenerator(Operand<TInt64> seed, Operand<TInt64> seed2, Operand<TBool> reshuffle) The AnonymousSeedGenerator operationComputes the "logical or" of elements across dimensions of a tensor.Creates a 1-dimensional operand that represents a new shape containing the dimensions of the operand representing a shape, followed by the dimensions of an operand representing a shape to append.<T extends TType>
ApplyAdadelta<T> TrainOps.applyAdadelta(Operand<T> var, Operand<T> accum, Operand<T> accumUpdate, Operand<T> lr, Operand<T> rho, Operand<T> epsilon, Operand<T> grad, ApplyAdadelta.Options... options) Update '*var' according to the adadelta scheme.<T extends TType>
ApplyAdagrad<T> TrainOps.applyAdagrad(Operand<T> var, Operand<T> accum, Operand<T> lr, Operand<T> grad, ApplyAdagrad.Options... options) Update '*var' according to the adagrad scheme.<T extends TType>
ApplyAdagradDa<T> TrainOps.applyAdagradDa(Operand<T> var, Operand<T> gradientAccumulator, Operand<T> gradientSquaredAccumulator, Operand<T> grad, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<TInt64> globalStep, ApplyAdagradDa.Options... options) Update '*var' according to the proximal adagrad scheme.<T extends TType>
ApplyAdagradV2<T> TrainOps.applyAdagradV2(Operand<T> var, Operand<T> accum, Operand<T> lr, Operand<T> epsilon, Operand<T> grad, ApplyAdagradV2.Options... options) Update '*var' according to the adagrad scheme.TrainOps.applyAdam(Operand<T> var, Operand<T> m, Operand<T> v, Operand<T> beta1Power, Operand<T> beta2Power, Operand<T> lr, Operand<T> beta1, Operand<T> beta2, Operand<T> epsilon, Operand<T> grad, ApplyAdam.Options... options) Update '*var' according to the Adam algorithm.<T extends TType>
ApplyAdaMax<T> TrainOps.applyAdaMax(Operand<T> var, Operand<T> m, Operand<T> v, Operand<T> beta1Power, Operand<T> lr, Operand<T> beta1, Operand<T> beta2, Operand<T> epsilon, Operand<T> grad, ApplyAdaMax.Options... options) Update '*var' according to the AdaMax algorithm.<T extends TType>
ApplyAddSign<T> TrainOps.applyAddSign(Operand<T> var, Operand<T> m, Operand<T> lr, Operand<T> alpha, Operand<T> signDecay, Operand<T> beta, Operand<T> grad, ApplyAddSign.Options... options) Update '*var' according to the AddSign update.<T extends TType>
ApplyCenteredRmsProp<T> TrainOps.applyCenteredRmsProp(Operand<T> var, Operand<T> mg, Operand<T> ms, Operand<T> mom, Operand<T> lr, Operand<T> rho, Operand<T> momentum, Operand<T> epsilon, Operand<T> grad, ApplyCenteredRmsProp.Options... options) Update '*var' according to the centered RMSProp algorithm.TrainOps.applyFtrl(Operand<T> var, Operand<T> accum, Operand<T> linear, Operand<T> grad, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<T> l2Shrinkage, Operand<T> lrPower, ApplyFtrl.Options... options) Update '*var' according to the Ftrl-proximal scheme.<T extends TType>
ApplyGradientDescent<T> TrainOps.applyGradientDescent(Operand<T> var, Operand<T> alpha, Operand<T> delta, ApplyGradientDescent.Options... options) Update '*var' by subtracting 'alpha' * 'delta' from it.<T extends TType>
ApplyMomentum<T> TrainOps.applyMomentum(Operand<T> var, Operand<T> accum, Operand<T> lr, Operand<T> grad, Operand<T> momentum, ApplyMomentum.Options... options) Update '*var' according to the momentum scheme.<T extends TType>
ApplyPowerSign<T> TrainOps.applyPowerSign(Operand<T> var, Operand<T> m, Operand<T> lr, Operand<T> logbase, Operand<T> signDecay, Operand<T> beta, Operand<T> grad, ApplyPowerSign.Options... options) Update '*var' according to the AddSign update.<T extends TType>
ApplyProximalAdagrad<T> TrainOps.applyProximalAdagrad(Operand<T> var, Operand<T> accum, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<T> grad, ApplyProximalAdagrad.Options... options) Update '*var' and '*accum' according to FOBOS with Adagrad learning rate.<T extends TType>
ApplyProximalGradientDescent<T> TrainOps.applyProximalGradientDescent(Operand<T> var, Operand<T> alpha, Operand<T> l1, Operand<T> l2, Operand<T> delta, ApplyProximalGradientDescent.Options... options) Update '*var' as FOBOS algorithm with fixed learning rate.<T extends TType>
ApplyRmsProp<T> TrainOps.applyRmsProp(Operand<T> var, Operand<T> ms, Operand<T> mom, Operand<T> lr, Operand<T> rho, Operand<T> momentum, Operand<T> epsilon, Operand<T> grad, ApplyRmsProp.Options... options) Update '*var' according to the RMSProp algorithm.<T extends TType>
ApproximateEqualMathOps.approximateEqual(Operand<T> x, Operand<T> y, ApproximateEqual.Options... options) Returns the truth value of abs(x-y) < tolerance element-wise.<T extends TNumber>
ApproxTopK<T> Ops.approxTopK(Operand<T> input, Long k, ApproxTopK.Options... options) Returns min/max k values and their indices of the input operand in an approximate manner.Returns the index with the largest value across dimensions of a tensor.MathOps.argMax(Operand<? extends TType> input, Operand<? extends TNumber> dimension, Class<V> outputType) Returns the index with the largest value across dimensions of a tensor.Returns the index with the smallest value across dimensions of a tensor.MathOps.argMin(Operand<? extends TType> input, Operand<? extends TNumber> dimension, Class<V> outputType) Returns the index with the smallest value across dimensions of a tensor.Computes the trignometric inverse sine of x element-wise.Computes inverse hyperbolic sine of x element-wise.DataOps.assertCardinalityDataset(Operand<? extends TType> inputDataset, Operand<TInt64> cardinality, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) The AssertCardinalityDataset operationDataExperimentalOps.assertNextDataset(Operand<? extends TType> inputDataset, Operand<TString> transformations, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) The ExperimentalAssertNextDataset operationDataOps.assertNextDataset(Operand<? extends TType> inputDataset, Operand<TString> transformations, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) A transformation that asserts which transformations happen next.DataOps.assertPrevDataset(Operand<? extends TType> inputDataset, Operand<TString> transformations, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) A transformation that asserts which transformations happened previously.Ops.assertThat(Operand<TBool> condition, Iterable<Operand<?>> data, AssertThat.Options... options) Asserts that the given condition is true.Ops.assign(Operand<T> ref, Operand<T> value, Assign.Options... options) Update 'ref' by assigning 'value' to it.Ops.assignAdd(Operand<T> ref, Operand<T> value, AssignAdd.Options... options) Update 'ref' by adding 'value' to it.Ops.assignAddVariableOp(Operand<? extends TType> resource, Operand<? extends TType> value) Adds a value to the current value of a variable.Ops.assignSub(Operand<T> ref, Operand<T> value, AssignSub.Options... options) Update 'ref' by subtracting 'value' from it.Ops.assignSubVariableOp(Operand<? extends TType> resource, Operand<? extends TType> value) Subtracts a value from the current value of a variable.XlaOps.assignVariableConcatND(Operand<? extends TType> resource, Iterable<Operand<? extends TType>> inputs, List<Long> numConcats, AssignVariableConcatND.Options... options) Concats input tensor across all dimensions.Ops.assignVariableOp(Operand<? extends TType> resource, Operand<? extends TType> value, AssignVariableOp.Options... options) Assigns a new value to a variable.DtypesOps.asString(Operand<? extends TType> input, AsString.Options... options) Converts each entry in the given tensor to strings.Computes the trignometric inverse tangent of x element-wise.Computes arctangent ofy/xelement-wise, respecting signs of the arguments.Computes inverse hyperbolic tangent of x element-wise.AudioOps.audioSpectrogram(Operand<TFloat32> input, Long windowSize, Long stride, AudioSpectrogram.Options... options) Produces a visualization of audio data over time.SummaryOps.audioSummary(Operand<TString> tag, Operand<TFloat32> tensor, Operand<TFloat32> sampleRate, AudioSummary.Options... options) Outputs aSummaryprotocol buffer with audio.DataExperimentalOps.autoShardDataset(Operand<? extends TType> inputDataset, Operand<TInt64> numWorkers, Operand<TInt64> index, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, AutoShardDataset.Options... options) Creates a dataset that shards the input dataset.DataOps.autoShardDataset(Operand<? extends TType> inputDataset, Operand<TInt64> numWorkers, Operand<TInt64> index, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, AutoShardDataset.Options... options) Creates a dataset that shards the input dataset.NnOps.avgPool(Operand<T> value, List<Long> ksize, List<Long> strides, String padding, AvgPool.Options... options) Performs average pooling on the input.NnOps.avgPool3d(Operand<T> input, List<Long> ksize, List<Long> strides, String padding, AvgPool3d.Options... options) Performs 3D average pooling on the input.<T extends TNumber>
AvgPool3dGrad<T> NnOps.avgPool3dGrad(Operand<TInt32> origInputShape, Operand<T> grad, List<Long> ksize, List<Long> strides, String padding, AvgPool3dGrad.Options... options) Computes gradients of average pooling function.<T extends TNumber>
AvgPoolGrad<T> NnOps.avgPoolGrad(Operand<TInt32> origInputShape, Operand<T> grad, List<Long> ksize, List<Long> strides, String padding, AvgPoolGrad.Options... options) Computes gradients of the average pooling function.<T extends TType>
BandedTriangularSolve<T> LinalgOps.bandedTriangularSolve(Operand<T> matrix, Operand<T> rhs, BandedTriangularSolve.Options... options) The BandedTriangularSolve operationCopy a tensor setting everything outside a central band in each innermost matrix to zero.Ops.barrierClose(Operand<TString> handle, BarrierClose.Options... options) Closes the given barrier.Ops.barrierIncompleteSize(Operand<TString> handle) Computes the number of incomplete elements in the given barrier.Ops.barrierInsertMany(Operand<TString> handle, Operand<TString> keys, Operand<? extends TType> values, Long componentIndex) For each key, assigns the respective value to the specified component.Ops.barrierReadySize(Operand<TString> handle) Computes the number of complete elements in the given barrier.Ops.barrierTakeMany(Operand<TString> handle, Operand<TInt32> numElements, List<Class<? extends TType>> componentTypes, BarrierTakeMany.Options... options) Takes the given number of completed elements from a barrier.<T extends TNumber>
BatchCholesky<T> LinalgOps.batchCholesky(Operand<T> input) The BatchCholesky operation<T extends TNumber>
BatchCholeskyGrad<T> LinalgOps.batchCholeskyGrad(Operand<T> l, Operand<T> grad) The BatchCholeskyGrad operationDataOps.batchDataset(Operand<? extends TType> inputDataset, Operand<TInt64> batchSize, Operand<TBool> dropRemainder, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, BatchDataset.Options... options) Creates a dataset that batchesbatch_sizeelements frominput_dataset.The BatchFFT operationSignalOps.batchFft2d(Operand<? extends TType> input) The BatchFFT2D operationSignalOps.batchFft3d(Operand<? extends TType> input) The BatchFFT3D operationThe BatchIFFT operationSignalOps.batchIfft2d(Operand<? extends TType> input) The BatchIFFT2D operationSignalOps.batchIfft3d(Operand<? extends TType> input) The BatchIFFT3D operation<V extends TType>
BatchMatMul<V> TrainOps.batchMatMul(Operand<? extends TType> x, Operand<? extends TType> y, Class<V> Tout, BatchMatMul.Options... options) Multiplies slices of two tensors in batches.<T extends TType>
BatchMatrixBandPart<T> The BatchMatrixBandPart operation<T extends TType>
BatchMatrixDeterminant<T> LinalgOps.batchMatrixDeterminant(Operand<T> input) The BatchMatrixDeterminant operation<T extends TType>
BatchMatrixDiag<T> LinalgOps.batchMatrixDiag(Operand<T> diagonal) The BatchMatrixDiag operation<T extends TType>
BatchMatrixDiagPart<T> LinalgOps.batchMatrixDiagPart(Operand<T> input) The BatchMatrixDiagPart operation<T extends TNumber>
BatchMatrixInverse<T> LinalgOps.batchMatrixInverse(Operand<T> input, BatchMatrixInverse.Options... options) The BatchMatrixInverse operation DEPRECATED: This operation is deprecated and will be removed in a future version.<T extends TType>
BatchMatrixSetDiag<T> LinalgOps.batchMatrixSetDiag(Operand<T> input, Operand<T> diagonal) The BatchMatrixSetDiag operation<T extends TNumber>
BatchMatrixSolve<T> LinalgOps.batchMatrixSolve(Operand<T> matrix, Operand<T> rhs, BatchMatrixSolve.Options... options) The BatchMatrixSolve operation<T extends TNumber>
BatchMatrixSolveLs<T> LinalgOps.batchMatrixSolveLs(Operand<T> matrix, Operand<T> rhs, Operand<TFloat64> l2Regularizer, BatchMatrixSolveLs.Options... options) The BatchMatrixSolveLs operation<T extends TNumber>
BatchMatrixTriangularSolve<T> LinalgOps.batchMatrixTriangularSolve(Operand<T> matrix, Operand<T> rhs, BatchMatrixTriangularSolve.Options... options) The BatchMatrixTriangularSolve operation<T extends TType>
BatchNormWithGlobalNormalization<T> NnOps.batchNormWithGlobalNormalization(Operand<T> t, Operand<T> m, Operand<T> v, Operand<T> beta, Operand<T> gamma, Float varianceEpsilon, Boolean scaleAfterNormalization) Batch normalization.<T extends TType>
BatchNormWithGlobalNormalizationGrad<T> NnOps.batchNormWithGlobalNormalizationGrad(Operand<T> t, Operand<T> m, Operand<T> v, Operand<T> gamma, Operand<T> backprop, Float varianceEpsilon, Boolean scaleAfterNormalization) Gradients for batch normalization.<T extends TNumber>
BatchSelfAdjointEig<T> LinalgOps.batchSelfAdjointEig(Operand<T> input, BatchSelfAdjointEig.Options... options) The BatchSelfAdjointEigV2 operationLinalgOps.batchSvd(Operand<T> input, BatchSvd.Options... options) The BatchSvd operation<T extends TType>
BatchToSpace<T> Ops.batchToSpace(Operand<T> input, Operand<? extends TNumber> crops, Long blockSize) BatchToSpace for 4-D tensors of type T.<T extends TType>
BatchToSpaceNd<T> Ops.batchToSpaceNd(Operand<T> input, Operand<? extends TNumber> blockShape, Operand<? extends TNumber> crops) BatchToSpace for N-D tensors of type T.The BesselI0 operationThe BesselI0e operationThe BesselI1 operationThe BesselI1e operationThe BesselJ0 operationThe BesselJ1 operationThe BesselK0 operationThe BesselK0e operationThe BesselK1 operationThe BesselK1e operationThe BesselY0 operationThe BesselY1 operationCompute the regularized incomplete beta integral \(I_x(a, b)\).NnOps.biasAdd(Operand<T> value, Operand<T> bias, BiasAdd.Options... options) Addsbiastovalue.<T extends TType>
BiasAddGrad<T> NnOps.biasAddGrad(Operand<T> outBackprop, BiasAddGrad.Options... options) The backward operation for "BiasAdd" on the "bias" tensor.Counts the number of occurrences of each value in an integer array.Bitcasts a tensor from one type to another without copying data.<T extends TNumber>
BitwiseAnd<T> BitwiseOps.bitwiseAnd(Operand<T> x, Operand<T> y) Elementwise computes the bitwise AND ofxandy.Elementwise computes the bitwise OR ofxandy.<T extends TNumber>
BitwiseXor<T> BitwiseOps.bitwiseXor(Operand<T> x, Operand<T> y) Elementwise computes the bitwise XOR ofxandy.NnOps.blockLSTM(Operand<TInt64> seqLenMax, Operand<T> x, Operand<T> csPrev, Operand<T> hPrev, Operand<T> w, Operand<T> wci, Operand<T> wcf, Operand<T> wco, Operand<T> b, BlockLSTM.Options... options) Computes the LSTM cell forward propagation for all the time steps.<T extends TNumber>
BlockLSTMGrad<T> NnOps.blockLSTMGrad(Operand<TInt64> seqLenMax, Operand<T> x, Operand<T> csPrev, Operand<T> hPrev, Operand<T> w, Operand<T> wci, Operand<T> wcf, Operand<T> wco, Operand<T> b, Operand<T> i, Operand<T> cs, Operand<T> f, Operand<T> o, Operand<T> ci, Operand<T> co, Operand<T> h, Operand<T> csGrad, Operand<T> hGrad, Boolean usePeephole) Computes the LSTM cell backward propagation for the entire time sequence.Ops.booleanMask(Operand<T> tensor, Operand<TBool> mask, BooleanMask.Options... options) Apply boolean mask to tensor.Ops.booleanMaskUpdate(Operand<T> tensor, Operand<TBool> mask, Operand<T> updates, BooleanMaskUpdate.Options... options) Updates a tensor at the masked values, and returns the updated tensor.<T extends TNumber>
BroadcastDynamicShape<T> Ops.broadcastDynamicShape(Operand<T> s0, Operand<T> s1) Return the shape of s0 op s1 with broadcast.<T extends TNumber>
BroadcastGradientArgs<T> Ops.broadcastGradientArgs(Operand<T> s0, Operand<T> s1) Return the reduction indices for computing gradients of s0 op s1 with broadcast.<T extends TType>
BroadcastTo<T> Ops.broadcastTo(Operand<T> input, Operand<? extends TNumber> shape) Broadcast an array for a compatible shape.Bucketizes 'input' based on 'boundaries'.DataExperimentalOps.bytesProducedStatsDataset(Operand<? extends TType> inputDataset, Operand<TString> tag, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Records the bytes size of each element ofinput_datasetin a StatsAggregator.DataOps.bytesProducedStatsDataset(Operand<? extends TType> inputDataset, Operand<TString> tag, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Records the bytes size of each element ofinput_datasetin a StatsAggregator.DataOps.cacheDataset(Operand<? extends TType> inputDataset, Operand<TString> filename, Operand<? extends TType> cache, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, CacheDataset.Options... options) The CacheDatasetV2 operationOperand<?> Ops.call(ConcreteFunction function, Operand<?> argument) Calls the function in an execution environment, adding its graph as a function if it isn't already present.Ops.caseOp(Operand<TInt32> branchIndex, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, List<ConcreteFunction> branches, Case.Options... options) An n-way switch statement which calls a single branch function.DtypesOps.cast(Operand<? extends TType> x, Class<U> DstT, Cast.Options... options) Cast x of type SrcT to y of DstT.Returns element-wise smallest integer not less than x.<T extends TNumber>
CheckNumerics<T> DebuggingOps.checkNumerics(Operand<T> tensor, String message) Checks a tensor for NaN, -Inf and +Inf values.<T extends TType>
CheckPinned<T> Ops.checkPinned(Operand<T> tensor) Checks whether a tensor is located in host memory pinned for GPU.Computes the Cholesky decomposition of one or more square matrices.<T extends TNumber>
CholeskyGrad<T> LinalgOps.choleskyGrad(Operand<T> l, Operand<T> grad) Computes the reverse mode backpropagated gradient of the Cholesky algorithm.DataOps.chooseFastestBranchDataset(Operand<? extends TType> inputDataset, Operand<TInt64> ratioNumerator, Operand<TInt64> ratioDenominator, Iterable<Operand<?>> otherArguments, Long numElementsPerBranch, List<ConcreteFunction> branches, List<Long> otherArgumentsLengths, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) The ChooseFastestBranchDataset operation<T extends TType>
ClipByValue<T> Ops.clipByValue(Operand<T> t, Operand<T> clipValueMin, Operand<T> clipValueMax) Clips tensor values to a specified min and max.SummaryOps.closeSummaryWriter(Operand<? extends TType> writer) The CloseSummaryWriter operation<T extends TNumber>
CollectiveAllToAll<T> CollectiveOps.collectiveAllToAll(Operand<T> input, Operand<? extends TType> communicator, Operand<TInt32> groupAssignment, CollectiveAllToAll.Options... options) Mutually exchanges multiple tensors of identical type and shape.CollectiveOps.collectiveAssignGroup(Operand<TInt32> groupAssignment, Operand<TInt32> deviceIndex, Operand<TInt32> baseKey) Assign group keys based on group assignment.<U extends TType>
CollectiveBcastRecv<U> CollectiveOps.collectiveBcastRecv(Operand<TInt32> groupSize, Operand<TInt32> groupKey, Operand<TInt32> instanceKey, Operand<? extends TNumber> shape, Class<U> T, CollectiveBcastRecv.Options... options) Receives a tensor value broadcast from another device.<T extends TType>
CollectiveBcastSend<T> CollectiveOps.collectiveBcastSend(Operand<T> input, Operand<TInt32> groupSize, Operand<TInt32> groupKey, Operand<TInt32> instanceKey, CollectiveBcastSend.Options... options) Broadcasts a tensor value to one or more other devices.<T extends TNumber>
CollectiveGather<T> CollectiveOps.collectiveGather(Operand<T> input, Operand<TInt32> groupSize, Operand<TInt32> groupKey, Operand<TInt32> instanceKey, Iterable<Operand<? extends TType>> orderingToken, CollectiveGather.Options... options) Mutually accumulates multiple tensors of identical type and shape.CollectiveOps.collectiveInitializeCommunicator(Operand<TInt32> groupKey, Operand<TInt32> rank, Operand<TInt32> groupSize, CollectiveInitializeCommunicator.Options... options) Initializes a group for collective operations.<T extends TType>
CollectivePermute<T> CollectiveOps.collectivePermute(Operand<T> input, Operand<TInt32> sourceTargetPairs) An Op to permute tensors across replicated TPU instances.<T extends TNumber>
CollectiveReduce<T> CollectiveOps.collectiveReduce(Operand<T> input, Operand<? extends TType> communicator, Operand<TInt32> groupAssignment, String reduction, CollectiveReduce.Options... options) Mutually reduces multiple tensors of identical type and shape.<T extends TNumber>
CollectiveReduceScatter<T> CollectiveOps.collectiveReduceScatter(Operand<T> input, Operand<TInt32> groupSize, Operand<TInt32> groupKey, Operand<TInt32> instanceKey, Iterable<Operand<? extends TType>> orderingToken, String mergeOp, String finalOp, CollectiveReduceScatter.Options... options) Mutually reduces multiple tensors of identical type and shape and scatters the result.ImageOps.combinedNonMaxSuppression(Operand<TFloat32> boxes, Operand<TFloat32> scores, Operand<TInt32> maxOutputSizePerClass, Operand<TInt32> maxTotalSize, Operand<TFloat32> iouThreshold, Operand<TFloat32> scoreThreshold, CombinedNonMaxSuppression.Options... options) Greedily selects a subset of bounding boxes in descending order of score, This operation performs non_max_suppression on the inputs per batch, across all classes.TpuOps.compileSucceededAssert(Operand<TString> compilationStatus) Asserts that compilation succeeded.Converts two real numbers to a complex number.MathOps.complexAbs(Operand<? extends TType> x) Computes the complex absolute value of a tensor.<U extends TNumber>
ComplexAbs<U> MathOps.complexAbs(Operand<? extends TType> x, Class<U> Tout) Computes the complex absolute value of a tensor.Ops.compositeTensorVariantToComponents(Operand<? extends TType> encoded, String metadata, List<Class<? extends TType>> Tcomponents) Decodes avariantscalar Tensor into anExtensionTypevalue.NnOps.computeAccidentalHits(Operand<TInt64> trueClasses, Operand<TInt64> sampledCandidates, Long numTrue, ComputeAccidentalHits.Options... options) Computes the ids of the positions in sampled_candidates that match true_labels.TrainOps.computeBatchSize(Operand<? extends TType> inputDataset) Computes the static batch size of a dataset sans partial batches.Concatenates tensors along one dimension.DataOps.concatenateDataset(Operand<? extends TType> inputDataset, Operand<? extends TType> anotherDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ConcatenateDataset.Options... options) Creates a dataset that concatenatesinput_datasetwithanother_dataset.<T extends TNumber>
ConcatOffset<T> Ops.concatOffset(Operand<TInt32> concatDim, Iterable<Operand<T>> shape) Computes offsets of concat inputs within its output.TpuOps.configureTPUEmbeddingHost(Operand<TString> commonConfig, Operand<TString> memoryConfig, String config) An op that configures the TPUEmbedding software on a host.TpuOps.configureTPUEmbeddingMemory(Operand<TString> commonConfig) An op that configures the TPUEmbedding software on a host.Returns the complex conjugate of a complex number.<T extends TType>
ConjugateTranspose<T> LinalgOps.conjugateTranspose(Operand<T> x, Operand<? extends TNumber> perm) Shuffle dimensions of x according to a permutation and conjugate the result.Ops.constantOfSameType(Operand<T> toMatch, Number number) Creates a scalar of the same type astoMatch, with the value ofnumber.Ops.consumeMutexLock(Operand<? extends TType> mutexLock) This op consumes a lock created byMutexLock.NnOps.conv(Operand<T> input, Operand<T> filter, List<Long> strides, String padding, Conv.Options... options) Computes a N-D convolution given (N+1+batch_dims)-Dinputand (N+2)-Dfiltertensors.NnOps.conv2d(Operand<T> input, Operand<T> filter, List<Long> strides, String padding, Conv2d.Options... options) Computes a 2-D convolution given 4-Dinputandfiltertensors.<T extends TNumber>
Conv2dBackpropFilter<T> NnOps.conv2dBackpropFilter(Operand<T> input, Operand<TInt32> filterSizes, Operand<T> outBackprop, List<Long> strides, String padding, Conv2dBackpropFilter.Options... options) Computes the gradients of convolution with respect to the filter.<T extends TNumber>
Conv2dBackpropInput<T> NnOps.conv2dBackpropInput(Operand<TInt32> inputSizes, Operand<T> filter, Operand<T> outBackprop, List<Long> strides, String padding, Conv2dBackpropInput.Options... options) Computes the gradients of convolution with respect to the input.NnOps.conv3d(Operand<T> input, Operand<T> filter, List<Long> strides, String padding, Conv3d.Options... options) Computes a 3-D convolution given 5-Dinputandfiltertensors.<T extends TNumber>
Conv3dBackpropFilter<T> NnOps.conv3dBackpropFilter(Operand<T> input, Operand<TInt32> filterSizes, Operand<T> outBackprop, List<Long> strides, String padding, Conv3dBackpropFilter.Options... options) Computes the gradients of 3-D convolution with respect to the filter.<U extends TNumber>
Conv3dBackpropInput<U> NnOps.conv3dBackpropInput(Operand<? extends TNumber> inputSizes, Operand<U> filter, Operand<U> outBackprop, List<Long> strides, String padding, Conv3dBackpropInput.Options... options) Computes the gradients of 3-D convolution with respect to the input.TpuOps.convertToCooTensor(Operand<TInt32> indicesOrRowSplits, Operand<TInt32> values, Operand<TFloat32> weights, Long sampleCount, String combiner) The ConvertToCooTensor operationSparseOps.convertToListOfSparseCoreCooTensors(Operand<TInt32> indicesOrRowSplits, Operand<TInt32> values, Operand<TFloat32> weights, Long sampleCount, Long numScPerChip, Long rowOffset, Long colOffset, Long colShift, Long numScShards, Long stackedTableSampleCount, String combiner) The ConvertToListOfSparseCoreCooTensors operationSparseOps.convertToSparseCoreCsrWrappedCooTensor(Iterable<Operand<TInt32>> sortedRowIdsList, Iterable<Operand<TInt32>> sortedColIdsList, Iterable<Operand<TFloat32>> sortedGainsList, Iterable<Operand<TInt32>> idCountsList, Operand<TInt64> splits, Long sampleCountPerSc, Long numReplica, Long maxMinibatchesPerSc, Long maxIdsPerChipPerSample, Long tableVocabSize, Long featureWidth, String tableName, Boolean allowIdDropping) The ConvertToSparseCoreCsrWrappedCooTensor operation<T extends TType>
CopyToMesh<T> Ops.copyToMesh(Operand<T> input, String mesh) The CopyToMesh operation<T extends TType>
CopyToMeshGrad<T> Ops.copyToMeshGrad(Operand<T> input, Operand<T> forwardInput) The CopyToMeshGrad operationComputes cos of x element-wise.Computes hyperbolic cosine of x element-wise.Increments 'ref' until it reaches 'limit'.SummaryOps.createSummaryDbWriter(Operand<? extends TType> writer, Operand<TString> dbUri, Operand<TString> experimentName, Operand<TString> runName, Operand<TString> userName) The CreateSummaryDbWriter operationSummaryOps.createSummaryFileWriter(Operand<? extends TType> writer, Operand<TString> logdir, Operand<TInt32> maxQueue, Operand<TInt32> flushMillis, Operand<TString> filenameSuffix) The CreateSummaryFileWriter operationImageOps.cropAndResize(Operand<? extends TNumber> image, Operand<TFloat32> boxes, Operand<TInt32> boxInd, Operand<TInt32> cropSize, CropAndResize.Options... options) Extracts crops from the input image tensor and resizes them.ImageOps.cropAndResizeGradBoxes(Operand<TFloat32> grads, Operand<? extends TNumber> image, Operand<TFloat32> boxes, Operand<TInt32> boxInd, CropAndResizeGradBoxes.Options... options) Computes the gradient of the crop_and_resize op wrt the input boxes tensor.<T extends TNumber>
CropAndResizeGradImage<T> ImageOps.cropAndResizeGradImage(Operand<TFloat32> grads, Operand<TFloat32> boxes, Operand<TInt32> boxInd, Operand<TInt32> imageSize, Class<T> T, CropAndResizeGradImage.Options... options) Computes the gradient of the crop_and_resize op wrt the input image tensor.Compute the pairwise cross product.<T extends TNumber>
CrossReplicaSum<T> TpuOps.crossReplicaSum(Operand<T> input, Operand<TInt32> groupAssignment) An Op to sum inputs across replicated TPU instances.<T extends TType>
CSRSparseMatrixComponents<T> LinalgSparseOps.cSRSparseMatrixComponents(Operand<? extends TType> csrSparseMatrix, Operand<TInt32> index, Class<T> type) Reads out the CSR components at batchindex.<T extends TType>
CSRSparseMatrixToDense<T> LinalgSparseOps.cSRSparseMatrixToDense(Operand<? extends TType> sparseInput, Class<T> type) Convert a (possibly batched) CSRSparseMatrix to dense.<T extends TType>
CSRSparseMatrixToSparseTensor<T> LinalgSparseOps.cSRSparseMatrixToSparseTensor(Operand<? extends TType> sparseMatrix, Class<T> type) Converts a (possibly batched) CSRSparesMatrix to a SparseTensor.DataExperimentalOps.cSVDataset(Operand<TString> filenames, Operand<TString> compressionType, Operand<TInt64> bufferSize, Operand<TBool> header, Operand<TString> fieldDelim, Operand<TBool> useQuoteDelim, Operand<TString> naValue, Operand<TInt64> selectCols, Iterable<Operand<?>> recordDefaults, List<Shape> outputShapes) The ExperimentalCSVDataset operationDataOps.cSVDataset(Operand<TString> filenames, Operand<TString> compressionType, Operand<TInt64> bufferSize, Operand<TBool> header, Operand<TString> fieldDelim, Operand<TBool> useQuoteDelim, Operand<TString> naValue, Operand<TInt64> selectCols, Iterable<Operand<?>> recordDefaults, Operand<TInt64> excludeCols, List<Shape> outputShapes) The CSVDatasetV2 operation<T extends TNumber>
CtcBeamSearchDecoder<T> NnOps.ctcBeamSearchDecoder(Operand<T> inputs, Operand<TInt32> sequenceLength, Long beamWidth, Long topPaths, CtcBeamSearchDecoder.Options... options) Performs beam search decoding on the logits given in input.<T extends TNumber>
CtcGreedyDecoder<T> NnOps.ctcGreedyDecoder(Operand<T> inputs, Operand<TInt32> sequenceLength, CtcGreedyDecoder.Options... options) Performs greedy decoding on the logits given in inputs.NnOps.ctcLoss(Operand<T> inputs, Operand<TInt64> labelsIndices, Operand<TInt32> labelsValues, Operand<TInt32> sequenceLength, CtcLoss.Options... options) Calculates the CTC Loss (log probability) for each batch entry.NnOps.cTCLossV2(Operand<TFloat32> inputs, Operand<TInt64> labelsIndices, Operand<TInt32> labelsValues, Operand<TInt32> sequenceLength, CTCLossV2.Options... options) Calculates the CTC Loss (log probability) for each batch entry.NnOps.cudnnRNN(Operand<T> input, Operand<T> inputH, Operand<T> inputC, Operand<T> params, Operand<TInt32> sequenceLengths, CudnnRNN.Options... options) A RNN backed by cuDNN.<T extends TNumber>
CudnnRNNBackprop<T> NnOps.cudnnRNNBackprop(Operand<T> input, Operand<T> inputH, Operand<T> inputC, Operand<T> params, Operand<TInt32> sequenceLengths, Operand<T> output, Operand<T> outputH, Operand<T> outputC, Operand<T> outputBackprop, Operand<T> outputHBackprop, Operand<T> outputCBackprop, Operand<T> reserveSpace, Operand<? extends TType> hostReserved, CudnnRNNBackprop.Options... options) Backprop step of CudnnRNNV3.<T extends TNumber>
CudnnRNNCanonicalToParams<T> NnOps.cudnnRNNCanonicalToParams(Operand<TInt32> numLayers, Operand<TInt32> numUnits, Operand<TInt32> inputSize, Iterable<Operand<T>> weights, Iterable<Operand<T>> biases, CudnnRNNCanonicalToParams.Options... options) Converts CudnnRNN params from canonical form to usable form.<T extends TNumber, U extends TNumber>
CudnnRnnParamsSize<T> NnOps.cudnnRnnParamsSize(Operand<TInt32> numLayers, Operand<TInt32> numUnits, Operand<TInt32> inputSize, Class<U> T, Class<T> S, CudnnRnnParamsSize.Options... options) Computes size of weights that can be used by a Cudnn RNN model.<T extends TNumber>
CudnnRNNParamsToCanonical<T> NnOps.cudnnRNNParamsToCanonical(Operand<TInt32> numLayers, Operand<TInt32> numUnits, Operand<TInt32> inputSize, Operand<T> params, Long numParamsWeights, Long numParamsBiases, CudnnRNNParamsToCanonical.Options... options) Retrieves CudnnRNN params in canonical form.MathOps.cumprod(Operand<T> x, Operand<? extends TNumber> axis, Cumprod.Options... options) Compute the cumulative product of the tensorxalongaxis.MathOps.cumsum(Operand<T> x, Operand<? extends TNumber> axis, Cumsum.Options... options) Compute the cumulative sum of the tensorxalongaxis.<T extends TNumber>
CumulativeLogsumexp<T> MathOps.cumulativeLogsumexp(Operand<T> x, Operand<? extends TNumber> axis, CumulativeLogsumexp.Options... options) Compute the cumulative product of the tensorxalongaxis.<T extends TNumber>
DataFormatDimMap<T> NnOps.dataFormatDimMap(Operand<T> x, DataFormatDimMap.Options... options) Returns the dimension index in the destination data format given the one in the source data format.<T extends TNumber>
DataFormatVecPermute<T> NnOps.dataFormatVecPermute(Operand<T> x, DataFormatVecPermute.Options... options) Permute input tensor fromsrc_formattodst_format.DataOps.dataServiceDataset(Operand<TString> datasetId, Operand<TString> processingMode, Operand<TString> address, Operand<TString> protocol, Operand<TString> jobName, Operand<TInt64> consumerIndex, Operand<TInt64> numConsumers, Operand<TInt64> maxOutstandingRequests, Operand<? extends TType> iterationCounter, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ConcreteFunction uncompressFn, DataServiceDataset.Options... options) Creates a dataset that reads data from the tf.data service.DataExperimentalOps.datasetCardinality(Operand<? extends TType> inputDataset) Returns the cardinality ofinput_dataset.DataOps.datasetCardinality(Operand<? extends TType> inputDataset, DatasetCardinality.Options... options) Returns the cardinality ofinput_dataset.DataOps.datasetFingerprint(Operand<? extends TType> inputDataset) Returns the fingerprint ofinput_dataset.DataOps.datasetFromGraph(Operand<TString> graphDef) Creates a dataset from the givengraph_def.DataOps.datasetToGraph(Operand<? extends TType> inputDataset, DatasetToGraph.Options... options) Returns a serialized GraphDef representinginput_dataset.DataOps.datasetToSingleElement(Operand<? extends TType> dataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, DatasetToSingleElement.Options... options) Outputs the single element from the given dataset.DataOps.datasetToTfRecord(Operand<? extends TType> inputDataset, Operand<TString> filename, Operand<TString> compressionType) Writes the given dataset to the given file using the TFRecord format.DataExperimentalOps.datasetToTFRecord(Operand<? extends TType> inputDataset, Operand<TString> filename, Operand<TString> compressionType) Writes the given dataset to the given file using the TFRecord format.The Dawsn operationImageOps.decodeAndCropJpeg(Operand<TString> contents, Operand<TInt32> cropWindow, DecodeAndCropJpeg.Options... options) Decode and Crop a JPEG-encoded image to a uint8 tensor.IoOps.decodeBase64(Operand<TString> input) Decode web-safe base64-encoded strings.ImageOps.decodeBmp(Operand<TString> contents, DecodeBmp.Options... options) Decode the first frame of a BMP-encoded image to a uint8 tensor.IoOps.decodeCompressed(Operand<TString> bytes, DecodeCompressed.Options... options) Decompress strings.IoOps.decodeCsv(Operand<TString> records, Iterable<Operand<?>> recordDefaults, DecodeCsv.Options... options) Convert CSV records to tensors.Decode the frame(s) of a GIF-encoded image to a uint8 tensor.<T extends TNumber>
DecodeImage<T> ImageOps.decodeImage(Operand<TString> contents, Class<T> dtype, DecodeImage.Options... options) Function for decode_bmp, decode_gif, decode_jpeg, and decode_png.ImageOps.decodeImage(Operand<TString> contents, DecodeImage.Options[] options) Function for decode_bmp, decode_gif, decode_jpeg, and decode_png.ImageOps.decodeJpeg(Operand<TString> contents, DecodeJpeg.Options... options) Decode a JPEG-encoded image to a uint8 tensor.IoOps.decodeJsonExample(Operand<TString> jsonExamples) Convert JSON-encoded Example records to binary protocol buffer strings.<T extends TNumber>
DecodePaddedRaw<T> IoOps.decodePaddedRaw(Operand<TString> inputBytes, Operand<TInt32> fixedLength, Class<T> outType, DecodePaddedRaw.Options... options) Reinterpret the bytes of a string as a vector of numbers.ImageOps.decodePng(Operand<TString> contents, Class<T> dtype, DecodePng.Options... options) Decode a PNG-encoded image to a uint8 or uint16 tensor.ImageOps.decodePng(Operand<TString> contents, DecodePng.Options[] options) Decode a PNG-encoded image to a uint8 or uint16 tensor.Ops.decodeProto(Operand<TString> bytes, String messageType, List<String> fieldNames, List<Class<? extends TType>> outputTypes, DecodeProto.Options... options) The op extracts fields from a serialized protocol buffers message into tensors.IoOps.decodeRaw(Operand<TString> bytes, Class<T> outType, DecodeRaw.Options... options) Reinterpret the bytes of a string as a vector of numbers.AudioOps.decodeWav(Operand<TString> contents, DecodeWav.Options... options) Decode a 16-bit PCM WAV file to a float tensor.Makes a copy ofx.DataOps.deleteIterator(Operand<? extends TType> handle, Operand<? extends TType> deleter) A container for an iterator resource.DataOps.deleteMemoryCache(Operand<? extends TType> handle, Operand<? extends TType> deleter) The DeleteMemoryCache operationDataOps.deleteMultiDeviceIterator(Operand<? extends TType> multiDeviceIterator, Iterable<Operand<? extends TType>> iterators, Operand<? extends TType> deleter) A container for an iterator resource.RandomOps.deleteRandomSeedGenerator(Operand<? extends TType> handle, Operand<? extends TType> deleter) The DeleteRandomSeedGenerator operationRandomOps.deleteSeedGenerator(Operand<? extends TType> handle, Operand<? extends TType> deleter) The DeleteSeedGenerator operationOps.deleteSessionTensor(Operand<TString> handle) Delete the tensor specified by its handle in the session.<U extends TNumber, T extends TNumber>
DenseBincount<U> MathOps.denseBincount(Operand<T> input, Operand<T> sizeOutput, Operand<U> weights, DenseBincount.Options... options) Counts the number of occurrences of each value in an integer array.<U extends TNumber>
DenseCountSparseOutput<U> SparseOps.denseCountSparseOutput(Operand<? extends TNumber> values, Operand<U> weights, Boolean binaryOutput, DenseCountSparseOutput.Options... options) Performs sparse-output bin counting for a tf.tensor input.LinalgSparseOps.denseToCSRSparseMatrix(Operand<? extends TType> denseInput, Operand<TInt64> indices) Converts a dense tensor to a (possibly batched) CSRSparseMatrix.<T extends TType>
DenseToDenseSetOperation<T> SparseOps.denseToDenseSetOperation(Operand<T> set1, Operand<T> set2, String setOperation, DenseToDenseSetOperation.Options... options) Applies set operation along last dimension of 2Tensorinputs.DataExperimentalOps.denseToSparseBatchDataset(Operand<? extends TType> inputDataset, Operand<TInt64> batchSize, Operand<TInt64> rowShape, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that batches input elements into a SparseTensor.DataOps.denseToSparseBatchDataset(Operand<? extends TType> inputDataset, Operand<TInt64> batchSize, Operand<TInt64> rowShape, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that batches input elements into a SparseTensor.<T extends TType>
DenseToSparseSetOperation<T> SparseOps.denseToSparseSetOperation(Operand<T> set1, Operand<TInt64> set2Indices, Operand<T> set2Values, Operand<TInt64> set2Shape, String setOperation, DenseToSparseSetOperation.Options... options) Applies set operation along last dimension ofTensorandSparseTensor.<T extends TType>
DepthToSpace<T> NnOps.depthToSpace(Operand<T> input, Long blockSize, DepthToSpace.Options... options) DepthToSpace for tensors of type T.<T extends TNumber>
DepthwiseConv2dNative<T> NnOps.depthwiseConv2dNative(Operand<T> input, Operand<T> filter, List<Long> strides, String padding, DepthwiseConv2dNative.Options... options) Computes a 2-D depthwise convolution given 4-Dinputandfiltertensors.<T extends TNumber>
DepthwiseConv2dNativeBackpropFilter<T> NnOps.depthwiseConv2dNativeBackpropFilter(Operand<T> input, Operand<TInt32> filterSizes, Operand<T> outBackprop, List<Long> strides, String padding, DepthwiseConv2dNativeBackpropFilter.Options... options) Computes the gradients of depthwise convolution with respect to the filter.<T extends TNumber>
DepthwiseConv2dNativeBackpropInput<T> NnOps.depthwiseConv2dNativeBackpropInput(Operand<TInt32> inputSizes, Operand<T> filter, Operand<T> outBackprop, List<Long> strides, String padding, DepthwiseConv2dNativeBackpropInput.Options... options) Computes the gradients of depthwise convolution with respect to the input.<U extends TNumber>
Dequantize<U> QuantizationOps.dequantize(Operand<? extends TNumber> input, Operand<TFloat32> minRange, Operand<TFloat32> maxRange, Class<U> dtype, Dequantize.Options... options) Dequantize the 'input' tensor into a float or bfloat16 Tensor.QuantizationOps.dequantize(Operand<? extends TNumber> input, Operand<TFloat32> minRange, Operand<TFloat32> maxRange, Dequantize.Options[] options) Dequantize the 'input' tensor into a float or bfloat16 Tensor.DataOps.deserializeIterator(Operand<? extends TType> resourceHandle, Operand<? extends TType> serialized) Converts the given variant tensor to an iterator and stores it in the given resource.<T extends TType>
DeserializeManySparse<T> IoOps.deserializeManySparse(Operand<TString> serializedSparse, Class<T> dtype) Deserialize and concatenateSparseTensorsfrom a serialized minibatch.<U extends TType>
DeserializeSparse<U> SparseOps.deserializeSparse(Operand<? extends TType> serializedSparse, Class<U> dtype) DeserializeSparseTensorobjects.Ops.destroyResourceOp(Operand<? extends TType> resource, DestroyResourceOp.Options... options) Deletes the resource specified by the handle.<T extends TType>
DestroyTemporaryVariable<T> Ops.destroyTemporaryVariable(Operand<T> ref, String varName) Destroys the temporary variable and returns its final value.Computes the determinant of one or more square matrices.Computes Psi, the derivative of Lgamma (the log of the absolute value ofGamma(x)), element-wise.<T extends TNumber>
Dilation2d<T> NnOps.dilation2d(Operand<T> input, Operand<T> filter, List<Long> strides, List<Long> rates, String padding) Computes the grayscale dilation of 4-Dinputand 3-Dfiltertensors.<T extends TNumber>
Dilation2dBackpropFilter<T> NnOps.dilation2dBackpropFilter(Operand<T> input, Operand<T> filter, Operand<T> outBackprop, List<Long> strides, List<Long> rates, String padding) Computes the gradient of morphological 2-D dilation with respect to the filter.<T extends TNumber>
Dilation2dBackpropInput<T> NnOps.dilation2dBackpropInput(Operand<T> input, Operand<T> filter, Operand<T> outBackprop, List<Long> strides, List<Long> rates, String padding) Computes the gradient of morphological 2-D dilation with respect to the input.DataExperimentalOps.directedInterleaveDataset(Operand<? extends TType> selectorInputDataset, Iterable<Operand<? extends TType>> dataInputDatasets, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) A substitute forInterleaveDataseton a fixed list ofNdatasets.DataOps.directedInterleaveDataset(Operand<? extends TType> selectorInputDataset, Iterable<Operand<? extends TType>> dataInputDatasets, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, DirectedInterleaveDataset.Options... options) A substitute forInterleaveDataseton a fixed list ofNdatasets.IoOps.disableCopyOnRead(Operand<? extends TType> resource) Turns off the copy-on-read mode.TrainOps.distributedSave(Operand<? extends TType> dataset, Operand<TString> directory, Operand<TString> address, DistributedSave.Options... options) The DistributedSave operationReturns x / y element-wise.Returns 0 if the denominator is zero.<T extends TNumber>
DrawBoundingBoxes<T> Draw bounding boxes on a batch of images.TpuOps.dTensorRestore(Operand<TString> prefix, Operand<TString> tensorNames, Operand<TString> shapeAndSlices, List<Shape> inputShapes, List<String> inputLayouts, List<Class<? extends TType>> dtypes) The DTensorRestoreV2 operationTpuOps.dynamicEnqueueTPUEmbeddingArbitraryTensorBatch(Iterable<Operand<? extends TNumber>> sampleIndicesOrRowSplits, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, Operand<TInt32> deviceOrdinal, DynamicEnqueueTPUEmbeddingArbitraryTensorBatch.Options... options) Eases the porting of code that uses tf.nn.embedding_lookup_sparse().TpuOps.dynamicEnqueueTPUEmbeddingRaggedTensorBatch(Iterable<Operand<? extends TNumber>> sampleSplits, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, Operand<TInt32> deviceOrdinal, List<Long> tableIds, DynamicEnqueueTPUEmbeddingRaggedTensorBatch.Options... options) The DynamicEnqueueTPUEmbeddingRaggedTensorBatch operation<T extends TType>
DynamicPartition<T> Ops.dynamicPartition(Operand<T> data, Operand<TInt32> partitions, Long numPartitions) Partitionsdataintonum_partitionstensors using indices frompartitions.<T extends TType>
EditDistanceOps.editDistance(Operand<TInt64> hypothesisIndices, Operand<T> hypothesisValues, Operand<TInt64> hypothesisShape, Operand<TInt64> truthIndices, Operand<T> truthValues, Operand<TInt64> truthShape, EditDistance.Options... options) Computes the (possibly normalized) Levenshtein Edit Distance.LinalgOps.eig(Operand<? extends TType> input, Class<U> Tout, Eig.Options... options) Computes the eigen decomposition of one or more square matrices.Computes the exponential linear function.Computes gradients for the exponential linear (Elu) operation.TpuOps.embeddingActivations(Operand<TFloat32> embeddingVariable, Operand<TFloat32> slicedActivations, Long tableId, Long lookupId) An op enabling differentiation of TPU Embeddings.Ops.empty(Operand<TInt32> shape, Class<T> dtype, Empty.Options... options) Creates a tensor with the given shape.<U extends TType>
EmptyTensorListOps.emptyTensorList(Operand<? extends TNumber> elementShape, Operand<TInt32> maxNumElements, Class<U> elementDtype) Creates and returns an empty tensor list.IoOps.encodeBase64(Operand<TString> input, EncodeBase64.Options... options) Encode strings into web-safe base64 format.ImageOps.encodeJpeg(Operand<TUint8> image, EncodeJpeg.Options... options) JPEG-encode an image.ImageOps.encodeJpegVariableQuality(Operand<TUint8> images, Operand<TInt32> quality) JPEG encode input image with provided compression quality.ImageOps.encodePng(Operand<? extends TNumber> image, EncodePng.Options... options) PNG-encode an image.Ops.encodeProto(Operand<TInt32> sizes, Iterable<Operand<?>> values, List<String> fieldNames, String messageType, EncodeProto.Options... options) The op serializes protobuf messages provided in the input tensors.Encode audio data using the WAV file format.TpuOps.enqueueTPUEmbeddingArbitraryTensorBatch(Iterable<Operand<? extends TNumber>> sampleIndicesOrRowSplits, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, EnqueueTPUEmbeddingArbitraryTensorBatch.Options... options) Eases the porting of code that uses tf.nn.embedding_lookup_sparse().TpuOps.enqueueTPUEmbeddingBatch(Iterable<Operand<TString>> batch, Operand<TString> modeOverride, EnqueueTPUEmbeddingBatch.Options... options) An op that enqueues a list of input batch tensors to TPUEmbedding.TpuOps.enqueueTPUEmbeddingIntegerBatch(Iterable<Operand<TInt32>> batch, Operand<TString> modeOverride, EnqueueTPUEmbeddingIntegerBatch.Options... options) An op that enqueues a list of input batch tensors to TPUEmbedding.TpuOps.enqueueTPUEmbeddingRaggedTensorBatch(Iterable<Operand<? extends TNumber>> sampleSplits, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, List<Long> tableIds, EnqueueTPUEmbeddingRaggedTensorBatch.Options... options) Eases the porting of code that uses tf.nn.embedding_lookup().TpuOps.enqueueTPUEmbeddingSparseBatch(Iterable<Operand<? extends TNumber>> sampleIndices, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, EnqueueTPUEmbeddingSparseBatch.Options... options) An op that enqueues TPUEmbedding input indices from a SparseTensor.TpuOps.enqueueTPUEmbeddingSparseTensorBatch(Iterable<Operand<? extends TNumber>> sampleIndices, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, List<Long> tableIds, EnqueueTPUEmbeddingSparseTensorBatch.Options... options) Eases the porting of code that uses tf.nn.embedding_lookup_sparse().<T extends TType>
EnsureShape<T> Ops.ensureShape(Operand<T> input, Shape shape) Ensures that the tensor's shape matches the expected shape.Ops.enter(Operand<T> data, String frameName, Enter.Options... options) Creates or finds a child frame, and makesdataavailable to the child frame.MathOps.equal(Operand<T> x, Operand<T> y, Equal.Options... options) Returns the truth value of (x == y) element-wise.Computes the Gauss error function ofxelement-wise.Computes the complementary error function ofxelement-wise.The Erfinv operation<T extends TType>
EuclideanNorm<T> LinalgOps.euclideanNorm(Operand<T> input, Operand<? extends TNumber> axis, EuclideanNorm.Options... options) Computes the euclidean norm of elements across dimensions of a tensor.TpuOps.execute(Iterable<Operand<?>> args, Operand<TString> key, List<Class<? extends TType>> Tresults) Op that loads and executes a TPU program on a TPU device.TpuOps.executeAndUpdateVariables(Iterable<Operand<?>> args, Operand<TString> key, List<Class<? extends TType>> Tresults, List<Long> deviceVarReadsIndices, List<Long> deviceVarUpdatesIndices) Op that executes a program with optional in-place variable updates.Exits the current frame to its parent frame.Computes exponential of x element-wise.<T extends TType>
ExpandDims<T> Ops.expandDims(Operand<T> input, Operand<? extends TNumber> axis) Inserts a dimension of 1 into a tensor's shape.The Expint operationComputesexp(x) - 1element-wise.ImageOps.extractGlimpse(Operand<TFloat32> input, Operand<TInt32> sizeOutput, Operand<TFloat32> offsets, ExtractGlimpse.Options... options) Extracts a glimpse from the input tensor.<T extends TType>
ExtractImagePatches<T> ImageOps.extractImagePatches(Operand<T> images, List<Long> ksizes, List<Long> strides, List<Long> rates, String padding) Extractpatchesfromimagesand put them in the "depth" output dimension.ImageOps.extractJpegShape(Operand<TString> contents) Extract the shape information of a JPEG-encoded image.<T extends TNumber>
ExtractJpegShape<T> ImageOps.extractJpegShape(Operand<TString> contents, Class<T> outputType) Extract the shape information of a JPEG-encoded image.<T extends TNumber>
ExtractVolumePatches<T> Extractpatchesfrominputand put them in the"depth"output dimension. 3D extension ofextract_image_patches.QuantizationOps.fakeQuantWithMinMaxArgs(Operand<TFloat32> inputs, FakeQuantWithMinMaxArgs.Options... options) Fake-quantize the 'inputs' tensor, type float to 'outputs' tensor of same shape and type.QuantizationOps.fakeQuantWithMinMaxArgsGradient(Operand<TFloat32> gradients, Operand<TFloat32> inputs, FakeQuantWithMinMaxArgsGradient.Options... options) Compute gradients for a FakeQuantWithMinMaxArgs operation.QuantizationOps.fakeQuantWithMinMaxVars(Operand<TFloat32> inputs, Operand<TFloat32> min, Operand<TFloat32> max, FakeQuantWithMinMaxVars.Options... options) Fake-quantize the 'inputs' tensor of type float via global float scalars Fake-quantize theinputstensor of type float via global float scalarsminandmaxtooutputstensor of same shape asinputs.QuantizationOps.fakeQuantWithMinMaxVarsGradient(Operand<TFloat32> gradients, Operand<TFloat32> inputs, Operand<TFloat32> min, Operand<TFloat32> max, FakeQuantWithMinMaxVarsGradient.Options... options) Compute gradients for a FakeQuantWithMinMaxVars operation.QuantizationOps.fakeQuantWithMinMaxVarsPerChannel(Operand<TFloat32> inputs, Operand<TFloat32> min, Operand<TFloat32> max, FakeQuantWithMinMaxVarsPerChannel.Options... options) Fake-quantize the 'inputs' tensor of type float via per-channel floats Fake-quantize theinputstensor of type float per-channel and one of the shapes:[d],[b, d][b, h, w, d]via per-channel floatsminandmaxof shape[d]tooutputstensor of same shape asinputs.QuantizationOps.fakeQuantWithMinMaxVarsPerChannelGradient(Operand<TFloat32> gradients, Operand<TFloat32> inputs, Operand<TFloat32> min, Operand<TFloat32> max, FakeQuantWithMinMaxVarsPerChannelGradient.Options... options) Compute gradients for a FakeQuantWithMinMaxVarsPerChannel operation.Deprecated.Fast Fourier transform.2D fast Fourier transform.3D fast Fourier transform.ND fast Fourier transform.Ops.fileSystemSetConfiguration(Operand<TString> scheme, Operand<TString> key, Operand<TString> value) Set configuration of the file system.Creates a tensor filled with a scalar value.DataOps.filterByLastComponentDataset(Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset containing elements of first component ofinput_datasethaving true in the last component.DataOps.filterDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction predicate, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, FilterDataset.Options... options) Creates a dataset containing elements ofinput_datasetmatchingpredicate.DataOps.finalizeDataset(Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, FinalizeDataset.Options... options) Creates a dataset by applyingtf.data.Optionstoinput_dataset.TpuOps.finalizeTPUEmbedding(Operand<TString> commonConfig, Operand<TString> memoryConfig) An op that finalizes the TPUEmbedding configuration.Ops.fingerprint(Operand<? extends TType> data, Operand<TString> method) Generates fingerprint values.DataOps.fixedLengthRecordDataset(Operand<TString> filenames, Operand<TInt64> headerBytes, Operand<TInt64> recordBytes, Operand<TInt64> footerBytes, Operand<TInt64> bufferSize, Operand<TString> compressionType, FixedLengthRecordDataset.Options... options) The FixedLengthRecordDatasetV2 operationNnOps.fixedUnigramCandidateSampler(Operand<TInt64> trueClasses, Long numTrue, Long numSampled, Boolean unique, Long rangeMax, FixedUnigramCandidateSampler.Options... options) Generates labels for candidate sampling with a learned unigram distribution.DataOps.flatMapDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, FlatMapDataset.Options... options) Creates a dataset that appliesfto the outputs ofinput_dataset.Flatten the operand to 1 dimension.Flatten the operand to 1 dimensionReturns element-wise largest integer not greater than x.Returns x // y element-wise.Returns element-wise remainder of division.SummaryOps.flushSummaryWriter(Operand<? extends TType> writer) The FlushSummaryWriter operationOps.forOp(Operand<TInt32> start, Operand<TInt32> limit, Operand<TInt32> delta, Iterable<Operand<?>> input, ConcreteFunction body) Applies a for loop.<T extends TNumber>
FractionalAvgPool<T> NnOps.fractionalAvgPool(Operand<T> value, List<Float> poolingRatio, FractionalAvgPool.Options... options) Performs fractional average pooling on the input.<T extends TNumber>
FractionalAvgPoolGrad<T> NnOps.fractionalAvgPoolGrad(Operand<TInt64> origInputTensorShape, Operand<T> outBackprop, Operand<TInt64> rowPoolingSequence, Operand<TInt64> colPoolingSequence, FractionalAvgPoolGrad.Options... options) Computes gradient of the FractionalAvgPool function.<T extends TNumber>
FractionalMaxPool<T> NnOps.fractionalMaxPool(Operand<T> value, List<Float> poolingRatio, FractionalMaxPool.Options... options) Performs fractional max pooling on the input.<T extends TNumber>
FractionalMaxPoolGrad<T> NnOps.fractionalMaxPoolGrad(Operand<T> origInput, Operand<T> origOutput, Operand<T> outBackprop, Operand<TInt64> rowPoolingSequence, Operand<TInt64> colPoolingSequence, FractionalMaxPoolGrad.Options... options) Computes gradient of the FractionalMaxPool function.<T extends TNumber>
FresnelCos<T> MathSpecialOps.fresnelCos(Operand<T> x) The FresnelCos operation<T extends TNumber>
FresnelSin<T> MathSpecialOps.fresnelSin(Operand<T> x) The FresnelSin operation<T extends TNumber, U extends TNumber>
FusedBatchNorm<T, U> NnOps.fusedBatchNorm(Operand<T> x, Operand<U> scale, Operand<U> offset, Operand<U> mean, Operand<U> variance, FusedBatchNorm.Options... options) Batch normalization.<T extends TNumber, U extends TNumber>
FusedBatchNormGrad<T, U> NnOps.fusedBatchNormGrad(Operand<T> yBackprop, Operand<T> x, Operand<TFloat32> scale, Operand<U> reserveSpace1, Operand<U> reserveSpace2, Operand<U> reserveSpace3, FusedBatchNormGrad.Options... options) Gradient for batch normalization.<T extends TNumber>
FusedPadConv2d<T> NnOps.fusedPadConv2d(Operand<T> input, Operand<TInt32> paddings, Operand<T> filter, String mode, List<Long> strides, String padding) Performs a padding as a preprocess during a convolution.<T extends TNumber>
FusedResizeAndPadConv2d<T> NnOps.fusedResizeAndPadConv2d(Operand<T> input, Operand<TInt32> sizeOutput, Operand<TInt32> paddings, Operand<T> filter, String mode, List<Long> strides, String padding, FusedResizeAndPadConv2d.Options... options) Performs a resize and padding as a preprocess during a convolution.Ops.gather(Operand<T> params, Operand<? extends TNumber> indices, Operand<? extends TNumber> axis, Gather.Options... options) Gather slices fromparamsaxisaxisaccording toindices.Ops.gatherNd(Operand<T> params, Operand<? extends TNumber> indices, GatherNd.Options... options) Gather slices fromparamsinto a Tensor with shape specified byindices.ImageOps.generateBoundingBoxProposals(Operand<TFloat32> scores, Operand<TFloat32> bboxDeltas, Operand<TFloat32> imageInfo, Operand<TFloat32> anchors, Operand<TFloat32> nmsThreshold, Operand<TInt32> preNmsTopn, Operand<TFloat32> minSize, GenerateBoundingBoxProposals.Options... options) This op produces Region of Interests from given bounding boxes(bbox_deltas) encoded wrt anchors according to eq.2 in arXiv:1506.01497TrainOps.generateVocabRemapping(Operand<TString> newVocabFile, Operand<TString> oldVocabFile, Long newVocabOffset, Long numNewVocab, GenerateVocabRemapping.Options... options) Given a path to new and old vocabulary files, returns a remapping Tensor of lengthnum_new_vocab, whereremapping[i]contains the row number in the old vocabulary that corresponds to rowiin the new vocabulary (starting at linenew_vocab_offsetand up tonum_new_vocabentities), or-1if entryiin the new vocabulary is not in the old vocabulary.Ops.getElementAtIndex(Operand<? extends TType> dataset, Operand<TInt64> index, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Gets the element at the specified index in a dataset.TpuOps.getMinibatchesInCsrWithPhysicalReplica(Operand<TString> programKey, Operand<TInt32> rowIds, Operand<TInt32> colIds, Operand<TFloat32> gains, Operand<TInt64> splits, Operand<TInt32> idCounts, Long sampleCount, Long numReplica, Long maxMinibatchesPerSc, Long maxIdsPerChipPerSample, Long tableVocabSize, Long featureWidth, Long numScPerChip, String tableName, String miniBatchInCsr) The GetMinibatchesInCsrWithPhysicalReplica operationTpuOps.getMinibatchSplitsWithPhysicalReplica(Operand<TString> programKey, Operand<TInt32> rowIds, Operand<TInt32> colIds, Operand<TFloat32> gains, Long sampleCount, Long numReplica, Long tableVocabSize, Long featureWidth, Long numScPerChip, String tableName, String miniBatchSplits) The GetMinibatchSplitsWithPhysicalReplica operationOps.getOptions(Operand<? extends TType> inputDataset) Returns thetf.data.Optionsattached toinput_dataset.Ops.getSessionHandle(Operand<? extends TType> value) Store the input tensor in the state of the current session.<T extends TType>
GetSessionTensor<T> Ops.getSessionTensor(Operand<TString> handle, Class<T> dtype) Get the value of the tensor specified by its handle.Ops.gradients(Operand<?> y, Iterable<? extends Operand<?>> x, Gradients.Options... options) Adds operations to compute the partial derivatives of sum ofys w.r.txs, i.e.,d(y_1 + y_2 + ...)/dx_1, d(y_1 + y_2 + ...)/dx_2...Returns the truth value of (x > y) element-wise.<T extends TNumber>
GreaterEqualMathOps.greaterEqual(Operand<T> x, Operand<T> y) Returns the truth value of (x >= y) element-wise.DataExperimentalOps.groupByReducerDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> keyFuncOtherArguments, Iterable<Operand<?>> initFuncOtherArguments, Iterable<Operand<?>> reduceFuncOtherArguments, Iterable<Operand<?>> finalizeFuncOtherArguments, ConcreteFunction keyFunc, ConcreteFunction initFunc, ConcreteFunction reduceFunc, ConcreteFunction finalizeFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that computes a group-by oninput_dataset.DataOps.groupByReducerDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> keyFuncOtherArguments, Iterable<Operand<?>> initFuncOtherArguments, Iterable<Operand<?>> reduceFuncOtherArguments, Iterable<Operand<?>> finalizeFuncOtherArguments, ConcreteFunction keyFunc, ConcreteFunction initFunc, ConcreteFunction reduceFunc, ConcreteFunction finalizeFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that computes a group-by oninput_dataset.DataExperimentalOps.groupByWindowDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> keyFuncOtherArguments, Iterable<Operand<?>> reduceFuncOtherArguments, Iterable<Operand<?>> windowSizeFuncOtherArguments, ConcreteFunction keyFunc, ConcreteFunction reduceFunc, ConcreteFunction windowSizeFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that computes a windowed group-by oninput_dataset. // TODO(mrry): Support non-int64 keys.DataOps.groupByWindowDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> keyFuncOtherArguments, Iterable<Operand<?>> reduceFuncOtherArguments, Iterable<Operand<?>> windowSizeFuncOtherArguments, ConcreteFunction keyFunc, ConcreteFunction reduceFunc, ConcreteFunction windowSizeFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, GroupByWindowDataset.Options... options) Creates a dataset that computes a windowed group-by oninput_dataset. // TODO(mrry): Support non-int64 keys.<T extends TNumber>
GRUBlockCell<T> NnOps.gRUBlockCell(Operand<T> x, Operand<T> hPrev, Operand<T> wRu, Operand<T> wC, Operand<T> bRu, Operand<T> bC) Computes the GRU cell forward propagation for 1 time step.<T extends TNumber>
GRUBlockCellGrad<T> NnOps.gRUBlockCellGrad(Operand<T> x, Operand<T> hPrev, Operand<T> wRu, Operand<T> wC, Operand<T> bRu, Operand<T> bC, Operand<T> r, Operand<T> u, Operand<T> c, Operand<T> dH) Computes the GRU cell back-propagation for 1 time step.<T extends TType>
GuaranteeConst<T> Ops.guaranteeConst(Operand<T> input) Gives a guarantee to the TF runtime that the input tensor is a constant.<T extends TNumber>
HistogramFixedWidth<TInt32> Ops.histogramFixedWidth(Operand<T> values, Operand<T> valueRange, Operand<TInt32> nbins) Return histogram of values.<U extends TNumber, T extends TNumber>
HistogramFixedWidth<U> Ops.histogramFixedWidth(Operand<T> values, Operand<T> valueRange, Operand<TInt32> nbins, Class<U> dtype) Return histogram of values.SummaryOps.histogramSummary(Operand<TString> tag, Operand<? extends TNumber> values) Outputs aSummaryprotocol buffer with a histogram.Convert one or more images from HSV to RGB.Return a tensor with the same shape and contents as the input tensor or value.Inverse fast Fourier transform.Inverse 2D fast Fourier transform.Inverse 3D fast Fourier transform.ND inverse fast Fourier transform.Ops.ifOp(Operand<? extends TType> cond, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, ConcreteFunction thenBranch, ConcreteFunction elseBranch, If.Options... options) output = cond ?Compute the lower regularized incomplete Gamma functionP(a, x).Compute the upper regularized incomplete Gamma functionQ(a, x).<T extends TNumber>
IgammaGradA<T> MathOps.igammaGradA(Operand<T> a, Operand<T> x) Computes the gradient ofigamma(a, x)wrta.DataExperimentalOps.ignoreErrorsDataset(Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, IgnoreErrorsDataset.Options... options) Creates a dataset that contains the elements ofinput_datasetignoring errors.DataOps.ignoreErrorsDataset(Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, IgnoreErrorsDataset.Options... options) Creates a dataset that contains the elements ofinput_datasetignoring errors.Returns the imaginary part of a complex number.Returns the imaginary part of a complex number.<T extends TNumber>
ImageProjectiveTransformV2<T> ImageOps.imageProjectiveTransformV2(Operand<T> images, Operand<TFloat32> transforms, Operand<TInt32> outputShape, String interpolation, ImageProjectiveTransformV2.Options... options) Applies the given transform to each of the images.<T extends TNumber>
ImageProjectiveTransformV3<T> ImageOps.imageProjectiveTransformV3(Operand<T> images, Operand<TFloat32> transforms, Operand<TInt32> outputShape, Operand<TFloat32> fillValue, String interpolation, ImageProjectiveTransformV3.Options... options) Applies the given transform to each of the images.SummaryOps.imageSummary(Operand<TString> tag, Operand<? extends TNumber> tensor, ImageSummary.Options... options) Outputs aSummaryprotocol buffer with images.SummaryOps.importEvent(Operand<? extends TType> writer, Operand<TString> event) The ImportEvent operationDataOps.indexFlatMapDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> mapFuncOtherArgs, Iterable<Operand<?>> indexMapFuncOtherArgs, Operand<TInt64> outputCardinality, ConcreteFunction mapFunc, ConcreteFunction indexMapFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, IndexFlatMapDataset.Options... options) The IndexFlatMapDataset operationTpuOps.infeedEnqueue(Operand<? extends TType> input, InfeedEnqueue.Options... options) An op which feeds a single Tensor value into the computation.TpuOps.infeedEnqueuePrelinearizedBuffer(Operand<? extends TType> input, InfeedEnqueuePrelinearizedBuffer.Options... options) An op which enqueues prelinearized buffer into TPU infeed.Ops.initializeTable(Operand<? extends TType> tableHandle, Operand<? extends TType> keys, Operand<? extends TType> values) Table initializer that takes two tensors for keys and values respectively.DataOps.initializeTableFromDataset(Operand<? extends TType> tableHandle, Operand<? extends TType> dataset) The InitializeTableFromDataset operationOps.initializeTableFromTextFile(Operand<? extends TType> tableHandle, Operand<TString> filename, Long keyIndex, Long valueIndex, InitializeTableFromTextFile.Options... options) Initializes a table from a text file.<T extends TType>
InplaceAdd<T> Ops.inplaceAdd(Operand<T> x, Operand<TInt32> i, Operand<T> v) Adds v into specified rows of x.<T extends TType>
InplaceSub<T> Ops.inplaceSub(Operand<T> x, Operand<TInt32> i, Operand<T> v) Subtracts `v` into specified rows of `x`.<T extends TType>
InplaceUpdate<T> Ops.inplaceUpdate(Operand<T> x, Operand<TInt32> i, Operand<T> v) Updates specified rows 'i' with values 'v'.DataOps.interleaveDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> cycleLength, Operand<TInt64> blockLength, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, InterleaveDataset.Options... options) Creates a dataset that appliesfto the outputs ofinput_dataset.Says whether the targets are in the topKpredictions.LinalgOps.inv(Operand<T> input, Inv.Options... options) Computes the inverse of one or more square invertible matrices or their adjoints (conjugate transposes).Invert (flip) each bit of supported types; for example, typeuint8value 01010101 becomes 10101010.<T extends TNumber>
InvertPermutation<T> MathOps.invertPermutation(Operand<T> x) Computes the inverse permutation of a tensor.Computes the gradient for the inverse ofxwrt its input.Inverse real-valued fast Fourier transform.Inverse real-valued fast Fourier transform.Inverse 2D real-valued fast Fourier transform.Inverse 2D real-valued fast Fourier transform.Inverse 3D real-valued fast Fourier transform.Inverse 3D real-valued fast Fourier transform.ND inverse real fast Fourier transform.SignalOps.irfftNd(Operand<? extends TType> input, Operand<TInt32> fftLength, Operand<TInt32> axes, Class<U> Treal) ND inverse real fast Fourier transform.Returns which elements of x are finite.Returns which elements of x are Inf.Returns which elements of x are NaN.NnOps.isotonicRegression(Operand<? extends TNumber> input) Solves a batch of isotonic regression problems.<U extends TNumber>
IsotonicRegression<U> NnOps.isotonicRegression(Operand<? extends TNumber> input, Class<U> outputDtype) Solves a batch of isotonic regression problems.Ops.isVariableInitialized(Operand<? extends TType> ref) Checks whether a tensor has been initialized.DataOps.iteratorFromStringHandle(Operand<TString> stringHandle, List<Class<? extends TType>> outputTypes, IteratorFromStringHandle.Options... options) The IteratorFromStringHandleV2 operationDataExperimentalOps.iteratorGetDevice(Operand<? extends TType> resource) Returns the name of the device on whichresourcehas been placed.DataOps.iteratorGetDevice(Operand<? extends TType> resource) Returns the name of the device on whichresourcehas been placed.DataOps.iteratorGetNext(Operand<? extends TType> iterator, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Gets the next output from the given iterator .DataOps.iteratorGetNextAsOptional(Operand<? extends TType> iterator, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Gets the next output from the given iterator as an Optional variant.DataOps.iteratorGetNextSync(Operand<? extends TType> iterator, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Gets the next output from the given iterator.DataOps.iteratorToStringHandle(Operand<? extends TType> resourceHandle) Converts the givenresource_handlerepresenting an iterator to a string.ClusterOps.kMC2ChainInitialization(Operand<TFloat32> distances, Operand<TInt64> seed) Returns the index of a data point that should be added to the seed set.ClusterOps.kmeansPlusPlusInitialization(Operand<TFloat32> points, Operand<TInt64> numToSample, Operand<TInt64> seed, Operand<TInt64> numRetriesPerSample) Selects num_to_sample rows of input using the KMeans++ criterion.Ops.kthOrderStatistic(Operand<TFloat32> input, Long k) Computes the Kth order statistic of a data set.L2 Loss.DataExperimentalOps.latencyStatsDataset(Operand<? extends TType> inputDataset, Operand<TString> tag, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Records the latency of producinginput_datasetelements in a StatsAggregator.DataOps.latencyStatsDataset(Operand<? extends TType> inputDataset, Operand<TString> tag, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Records the latency of producinginput_datasetelements in a StatsAggregator.NnOps.leakyRelu(Operand<T> features, LeakyRelu.Options... options) Computes rectified linear:max(features, features * alpha).<T extends TNumber>
LeakyReluGrad<T> DataOps.leakyReluGrad(Operand<T> gradients, Operand<T> features, LeakyReluGrad.Options... options) Computes rectified linear gradients for a LeakyRelu operation.NnOps.learnedUnigramCandidateSampler(Operand<TInt64> trueClasses, Long numTrue, Long numSampled, Boolean unique, Long rangeMax, LearnedUnigramCandidateSampler.Options... options) Generates labels for candidate sampling with a learned unigram distribution.Elementwise computes the bitwise left-shift ofxandy.DataOps.legacyParallelInterleaveDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> cycleLength, Operand<TInt64> blockLength, Operand<TInt64> bufferOutputElements, Operand<TInt64> prefetchInputElements, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, LegacyParallelInterleaveDataset.Options... options) Creates a dataset that appliesfto the outputs ofinput_dataset.Returns the truth value of (x < y) element-wise.Returns the truth value of (x <= y) element-wise.Computes the log of the absolute value ofGamma(x)element-wise.Generates values in an interval.DataOps.listSnapshotChunksDataset(Operand<TString> snapshotPath, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) The ListSnapshotChunksDataset operationDataExperimentalOps.lmdbDataset(Operand<TString> filenames, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) The ExperimentalLMDBDataset operationDataOps.lMDBDataset(Operand<TString> filenames, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that emits the key-value pairs in one or more LMDB files.LinalgOps.loadAndRemapMatrix(Operand<TString> ckptPath, Operand<TString> oldTensorName, Operand<TInt64> rowRemapping, Operand<TInt64> colRemapping, Operand<TFloat32> initializingValues, Long numRows, Long numCols, LoadAndRemapMatrix.Options... options) Loads a 2-D (matrix)Tensorwith nameold_tensor_namefrom the checkpoint atckpt_pathand potentially reorders its rows and columns using the specified remappings.DataOps.loadDataset(Operand<TString> path, Iterable<Operand<?>> readerFuncOtherArgs, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ConcreteFunction readerFunc, LoadDataset.Options... options) The LoadDataset operationTpuOps.loadTPUEmbeddingAdadeltaParameters(Operand<TFloat32> parameters, Operand<TFloat32> accumulators, Operand<TFloat32> updates, Long numShards, Long shardId, LoadTPUEmbeddingAdadeltaParameters.Options... options) Load Adadelta embedding parameters.TpuOps.loadTPUEmbeddingAdagradMomentumParameters(Operand<TFloat32> parameters, Operand<TFloat32> accumulators, Operand<TFloat32> momenta, Long numShards, Long shardId, LoadTPUEmbeddingAdagradMomentumParameters.Options... options) Load Adagrad Momentum embedding parameters.TpuOps.loadTPUEmbeddingAdagradParameters(Operand<TFloat32> parameters, Operand<TFloat32> accumulators, Long numShards, Long shardId, LoadTPUEmbeddingAdagradParameters.Options... options) Load Adagrad embedding parameters.TpuOps.loadTPUEmbeddingADAMParameters(Operand<TFloat32> parameters, Operand<TFloat32> momenta, Operand<TFloat32> velocities, Long numShards, Long shardId, LoadTPUEmbeddingADAMParameters.Options... options) Load ADAM embedding parameters.TpuOps.loadTPUEmbeddingCenteredRMSPropParameters(Operand<TFloat32> parameters, Operand<TFloat32> ms, Operand<TFloat32> mom, Operand<TFloat32> mg, Long numShards, Long shardId, LoadTPUEmbeddingCenteredRMSPropParameters.Options... options) Load centered RMSProp embedding parameters.TpuOps.loadTPUEmbeddingFrequencyEstimatorParameters(Operand<TFloat32> parameters, Operand<TFloat32> lastHitStep, Long numShards, Long shardId, LoadTPUEmbeddingFrequencyEstimatorParameters.Options... options) Load frequency estimator embedding parameters.TpuOps.loadTPUEmbeddingFTRLParameters(Operand<TFloat32> parameters, Operand<TFloat32> accumulators, Operand<TFloat32> linears, Long numShards, Long shardId, LoadTPUEmbeddingFTRLParameters.Options... options) Load FTRL embedding parameters.TpuOps.loadTPUEmbeddingMDLAdagradLightParameters(Operand<TFloat32> parameters, Operand<TFloat32> accumulators, Operand<TFloat32> weights, Operand<TFloat32> benefits, Long numShards, Long shardId, LoadTPUEmbeddingMDLAdagradLightParameters.Options... options) Load MDL Adagrad Light embedding parameters.TpuOps.loadTPUEmbeddingMomentumParameters(Operand<TFloat32> parameters, Operand<TFloat32> momenta, Long numShards, Long shardId, LoadTPUEmbeddingMomentumParameters.Options... options) Load Momentum embedding parameters.TpuOps.loadTPUEmbeddingProximalAdagradParameters(Operand<TFloat32> parameters, Operand<TFloat32> accumulators, Long numShards, Long shardId, LoadTPUEmbeddingProximalAdagradParameters.Options... options) Load proximal Adagrad embedding parameters.TpuOps.loadTPUEmbeddingProximalYogiParameters(Operand<TFloat32> parameters, Operand<TFloat32> v, Operand<TFloat32> m, Long numShards, Long shardId, LoadTPUEmbeddingProximalYogiParameters.Options... options) The LoadTPUEmbeddingProximalYogiParameters operationTpuOps.loadTPUEmbeddingRMSPropParameters(Operand<TFloat32> parameters, Operand<TFloat32> ms, Operand<TFloat32> mom, Long numShards, Long shardId, LoadTPUEmbeddingRMSPropParameters.Options... options) Load RMSProp embedding parameters.TpuOps.loadTPUEmbeddingStochasticGradientDescentParameters(Operand<TFloat32> parameters, Long numShards, Long shardId, LoadTPUEmbeddingStochasticGradientDescentParameters.Options... options) Load SGD embedding parameters.<T extends TNumber>
LocalResponseNormalization<T> NnOps.localResponseNormalization(Operand<T> input, LocalResponseNormalization.Options... options) Local Response Normalization.<T extends TNumber>
LocalResponseNormalizationGrad<T> NnOps.localResponseNormalizationGrad(Operand<T> inputGrads, Operand<T> inputImage, Operand<T> outputImage, LocalResponseNormalizationGrad.Options... options) Gradients for Local Response Normalization.Computes natural logarithm of x element-wise.Computes natural logarithm of (1 + x) element-wise.MathOps.logicalAnd(Operand<TBool> x, Operand<TBool> y) Returns the truth value of x AND y element-wise.MathOps.logicalNot(Operand<TBool> x) Returns the truth value ofNOT xelement-wise.Returns the truth value of x OR y element-wise.<T extends TType>
LogMatrixDeterminant<T> LinalgOps.logMatrixDeterminant(Operand<T> input) Computes the sign and the log of the absolute value of the determinant of one or more square matrices.<T extends TNumber>
LogSoftmax<T> NnOps.logSoftmax(Operand<T> logits) Computes log softmax activations.RandomOps.logUniformCandidateSampler(Operand<TInt64> trueClasses, Long numTrue, Long numSampled, Boolean unique, Long rangeMax, LogUniformCandidateSampler.Options... options) Generates labels for candidate sampling with a log-uniform distribution.<T extends TType, U extends TType>
LookupTableExport<T, U> Ops.lookupTableExport(Operand<? extends TType> tableHandle, Class<T> Tkeys, Class<U> Tvalues) Outputs all keys and values in the table.<U extends TType>
LookupTableFind<U> Ops.lookupTableFind(Operand<? extends TType> tableHandle, Operand<? extends TType> keys, Operand<U> defaultValue) Looks up keys in a table, outputs the corresponding values.Ops.lookupTableImport(Operand<? extends TType> tableHandle, Operand<? extends TType> keys, Operand<? extends TType> values) Replaces the contents of the table with the specified keys and values.Ops.lookupTableInsert(Operand<? extends TType> tableHandle, Operand<? extends TType> keys, Operand<? extends TType> values) Updates the table to associates keys with values.Ops.lookupTableRemove(Operand<? extends TType> tableHandle, Operand<? extends TType> keys) Removes keys and its associated values from a table.Ops.lookupTableSize(Operand<? extends TType> tableHandle) Computes the number of elements in the given table.Forwards the input to the output.StringsOps.lower(Operand<TString> input, Lower.Options... options) Converts all uppercase characters into their respective lowercase replacements.<T extends TType>
LowerBound<TInt32> Ops.lowerBound(Operand<T> sortedInputs, Operand<T> values) Applies lower_bound(sorted_search_values, values) along each row.<U extends TNumber, T extends TType>
LowerBound<U> Ops.lowerBound(Operand<T> sortedInputs, Operand<T> values, Class<U> outType) Applies lower_bound(sorted_search_values, values) along each row.<T extends TNumber>
LSTMBlockCell<T> NnOps.lSTMBlockCell(Operand<T> x, Operand<T> csPrev, Operand<T> hPrev, Operand<T> w, Operand<T> wci, Operand<T> wcf, Operand<T> wco, Operand<T> b, LSTMBlockCell.Options... options) Computes the LSTM cell forward propagation for 1 time step.<T extends TNumber>
LSTMBlockCellGrad<T> NnOps.lSTMBlockCellGrad(Operand<T> x, Operand<T> csPrev, Operand<T> hPrev, Operand<T> w, Operand<T> wci, Operand<T> wcf, Operand<T> wco, Operand<T> b, Operand<T> i, Operand<T> cs, Operand<T> f, Operand<T> o, Operand<T> ci, Operand<T> co, Operand<T> csGrad, Operand<T> hGrad, Boolean usePeephole) Computes the LSTM cell backward propagation for 1 timestep.Computes the LU decomposition of one or more square matrices.Computes the LU decomposition of one or more square matrices.DataOps.makeIterator(Operand<? extends TType> dataset, Operand<? extends TType> iterator) Makes a new iterator from the givendatasetand stores it initerator.Ops.makeUnique(Operand<TFloat32> input) Make all elements in the non-Batch dimension unique, but "close" to their initial value.DataExperimentalOps.mapAndBatchDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> batchSize, Operand<TInt64> numParallelCalls, Operand<TBool> dropRemainder, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, MapAndBatchDataset.Options... options) Creates a dataset that fuses mapping with batching.DataOps.mapAndBatchDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> batchSize, Operand<TInt64> numParallelCalls, Operand<TBool> dropRemainder, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, MapAndBatchDataset.Options... options) Creates a dataset that fuses mapping with batching.DataExperimentalOps.mapDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, MapDataset.Options... options) Creates a dataset that appliesfto the outputs ofinput_dataset.DataOps.mapDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, MapDataset.Options... options) Creates a dataset that appliesfto the outputs ofinput_dataset.Ops.mapPeek(Operand<TInt64> key, Operand<TInt32> indices, List<Class<? extends TType>> dtypes, MapPeek.Options... options) Op peeks at the values at the specified key.Ops.mapStage(Operand<TInt64> key, Operand<TInt32> indices, Iterable<Operand<?>> values, List<Class<? extends TType>> dtypes, MapStage.Options... options) Stage (key, values) in the underlying container which behaves like a hashtable.Ops.mapUnstage(Operand<TInt64> key, Operand<TInt32> indices, List<Class<? extends TType>> dtypes, MapUnstage.Options... options) Op removes and returns the values associated with the key from the underlying container.Ops.mapUnstageNoKey(Operand<TInt32> indices, List<Class<? extends TType>> dtypes, MapUnstageNoKey.Options... options) Op removes and returns a random (key, value) from the underlying container.IoOps.matchingFiles(Operand<TString> pattern) Returns the set of files matching one or more glob patterns.DataExperimentalOps.matchingFilesDataset(Operand<TString> patterns) The ExperimentalMatchingFilesDataset operationDataOps.matchingFilesDataset(Operand<TString> patterns) The MatchingFilesDataset operationLinalgOps.matMul(Operand<T> a, Operand<T> b, MatMul.Options... options) Multiply the matrix "a" by the matrix "b".<T extends TType>
MatrixDiag<T> LinalgOps.matrixDiag(Operand<T> diagonal, Operand<TInt32> k, Operand<TInt32> numRows, Operand<TInt32> numCols, Operand<T> paddingValue) Returns a batched diagonal tensor with given batched diagonal values.<T extends TType>
MatrixDiagPart<T> LinalgOps.matrixDiagPart(Operand<T> input, Operand<TInt32> k, Operand<T> paddingValue) Returns the batched diagonal part of a batched tensor.<T extends TType>
MatrixDiagPartV3<T> LinalgOps.matrixDiagPartV3(Operand<T> input, Operand<TInt32> k, Operand<T> paddingValue, MatrixDiagPartV3.Options... options) Returns the batched diagonal part of a batched tensor.<T extends TType>
MatrixDiagV3<T> LinalgOps.matrixDiagV3(Operand<T> diagonal, Operand<TInt32> k, Operand<TInt32> numRows, Operand<TInt32> numCols, Operand<T> paddingValue, MatrixDiagV3.Options... options) Returns a batched diagonal tensor with given batched diagonal values.<T extends TType>
MatrixExponential<T> LinalgOps.matrixExponential(Operand<T> input) Deprecated, use python implementation tf.linalg.matrix_exponential.<T extends TType>
MatrixLogarithm<T> LinalgOps.matrixLogarithm(Operand<T> input) Computes the matrix logarithm of one or more square matrices: \(log(exp(A)) = A\)<T extends TType>
MatrixSetDiag<T> LinalgOps.matrixSetDiag(Operand<T> input, Operand<T> diagonal, Operand<TInt32> k, MatrixSetDiag.Options... options) Returns a batched matrix tensor with new batched diagonal values.<T extends TType>
MatrixSolveLs<T> LinalgOps.matrixSolveLs(Operand<T> matrix, Operand<T> rhs, Operand<TFloat64> l2Regularizer, MatrixSolveLs.Options... options) Solves one or more linear least-squares problems.Ops.max(Operand<T> input, Operand<? extends TNumber> axis, Max.Options... options) Computes the maximum of elements across dimensions of a tensor.Returns the max of x and y (i.e. x > y ?DataExperimentalOps.maxIntraOpParallelismDataset(Operand<? extends TType> inputDataset, Operand<TInt64> maxIntraOpParallelism, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that overrides the maximum intra-op parallelism.DataOps.maxIntraOpParallelismDataset(Operand<? extends TType> inputDataset, Operand<TInt64> maxIntraOpParallelism, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that overrides the maximum intra-op parallelism.NnOps.maxPool(Operand<T> input, Operand<TInt32> ksize, Operand<TInt32> strides, String padding, MaxPool.Options... options) Performs max pooling on the input.NnOps.maxPool3d(Operand<T> input, List<Long> ksize, List<Long> strides, String padding, MaxPool3d.Options... options) Performs 3D max pooling on the input.<U extends TNumber, T extends TNumber>
MaxPool3dGrad<U> NnOps.maxPool3dGrad(Operand<T> origInput, Operand<T> origOutput, Operand<U> grad, List<Long> ksize, List<Long> strides, String padding, MaxPool3dGrad.Options... options) Computes gradients of 3D max pooling function.<T extends TNumber>
MaxPool3dGradGrad<T> NnOps.maxPool3dGradGrad(Operand<T> origInput, Operand<T> origOutput, Operand<T> grad, List<Long> ksize, List<Long> strides, String padding, MaxPool3dGradGrad.Options... options) Computes second-order gradients of the maxpooling function.<T extends TNumber>
MaxPoolGrad<T> NnOps.maxPoolGrad(Operand<T> origInput, Operand<T> origOutput, Operand<T> grad, Operand<TInt32> ksize, Operand<TInt32> strides, String padding, MaxPoolGrad.Options... options) Computes gradients of the maxpooling function.<T extends TNumber>
MaxPoolGradGrad<T> NnOps.maxPoolGradGrad(Operand<T> origInput, Operand<T> origOutput, Operand<T> grad, Operand<TInt32> ksize, Operand<TInt32> strides, String padding, MaxPoolGradGrad.Options... options) Computes second-order gradients of the maxpooling function.<T extends TNumber>
MaxPoolGradGradWithArgmax<T> NnOps.maxPoolGradGradWithArgmax(Operand<T> input, Operand<T> grad, Operand<? extends TNumber> argmax, List<Long> ksize, List<Long> strides, String padding, MaxPoolGradGradWithArgmax.Options... options) Computes second-order gradients of the maxpooling function.<T extends TNumber>
MaxPoolGradWithArgmax<T> NnOps.maxPoolGradWithArgmax(Operand<T> input, Operand<T> grad, Operand<? extends TNumber> argmax, List<Long> ksize, List<Long> strides, String padding, MaxPoolGradWithArgmax.Options... options) Computes gradients of the maxpooling function.<T extends TNumber, U extends TNumber>
MaxPoolWithArgmax<T, U> NnOps.maxPoolWithArgmax(Operand<T> input, List<Long> ksize, List<Long> strides, Class<U> Targmax, String padding, MaxPoolWithArgmax.Options... options) Performs max pooling on the input and outputs both max values and indices.<T extends TNumber>
MaxPoolWithArgmax<T, TInt64> NnOps.maxPoolWithArgmax(Operand<T> input, List<Long> ksize, List<Long> strides, String padding, MaxPoolWithArgmax.Options[] options) Performs max pooling on the input and outputs both max values and indices.MathOps.mean(Operand<T> input, Operand<? extends TNumber> axis, Mean.Options... options) Computes the mean of elements across dimensions of a tensor.TpuOps.mergeDedupData(Operand<? extends TNumber> integerTensor, Operand<? extends TNumber> floatTensor, String tupleMask, MergeDedupData.Options... options) An op merges elements of integer and float tensors into deduplication data as XLA tuple.TrainOps.mergeV2Checkpoints(Operand<TString> checkpointPrefixes, Operand<TString> destinationPrefix, MergeV2Checkpoints.Options... options) V2 format specific: merges the metadata files of sharded checkpoints.Transforms a spectrogram into a form that's useful for speech recognition.Ops.min(Operand<T> input, Operand<? extends TNumber> axis, Min.Options... options) Computes the minimum of elements across dimensions of a tensor.Returns the min of x and y (i.e. x < y ?Pads a tensor with mirrored values.<T extends TType>
MirrorPadGrad<T> Ops.mirrorPadGrad(Operand<T> input, Operand<? extends TNumber> paddings, String mode) Gradient op forMirrorPadop.Returns element-wise remainder of division.DataOps.modelDataset(Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ModelDataset.Options... options) Identity transformation that models performance.Returns x * y element-wise.Returns x * y element-wise.DataOps.multiDeviceIteratorFromStringHandle(Operand<TString> stringHandle, List<Class<? extends TType>> outputTypes, MultiDeviceIteratorFromStringHandle.Options... options) Generates a MultiDeviceIterator resource from its provided string handle.DataOps.multiDeviceIteratorGetNextFromShard(Operand<? extends TType> multiDeviceIterator, Operand<TInt32> shardNum, Operand<TInt64> incarnationId, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Gets next element for the provided shard number.DataOps.multiDeviceIteratorInit(Operand<? extends TType> dataset, Operand<? extends TType> multiDeviceIterator, Operand<TInt64> maxBufferSize) Initializes the multi device iterator with the given dataset.DataOps.multiDeviceIteratorToStringHandle(Operand<? extends TType> multiDeviceIterator) Produces a string handle for the given MultiDeviceIterator.<U extends TNumber>
Multinomial<U> RandomOps.multinomial(Operand<? extends TNumber> logits, Operand<TInt32> numSamples, Class<U> outputDtype, Multinomial.Options... options) Draws samples from a multinomial distribution.RandomOps.multinomial(Operand<? extends TNumber> logits, Operand<TInt32> numSamples, Multinomial.Options[] options) Draws samples from a multinomial distribution.<T extends TType, U extends TType>
MutableDenseHashTableOps.mutableDenseHashTable(Operand<T> emptyKey, Operand<T> deletedKey, Class<U> valueDtype, MutableDenseHashTable.Options... options) Creates an empty hash table that uses tensors as the backing store.Locks a mutex resource.<T extends TNumber>
NcclAllReduce<T> DistributeOps.ncclAllReduce(Operand<T> input, String reduction, Long numDevices, String sharedName) Outputs a tensor containing the reduction across all input tensors.<T extends TNumber>
NcclAllReduce<T> Ops.ncclAllReduce(Operand<T> input, String reduction, Long numDevices, String sharedName) Deprecated.<T extends TNumber>
NcclBroadcast<T> DistributeOps.ncclBroadcast(Operand<T> input, Shape shape) Sendsinputto all devices that are connected to the output.<T extends TNumber>
NcclBroadcast<T> Ops.ncclBroadcast(Operand<T> input, Shape shape) Deprecated.useNcclBroadcastinsteadThe Ndtri operationSelects the k nearest centers for each point.Computes numerical negative value element-wise.TrainOps.negTrain(Operand<TFloat32> wIn, Operand<TFloat32> wOut, Operand<TInt32> examples, Operand<TInt32> labels, Operand<TFloat32> lr, List<Long> vocabCount, Long numNegativeSamples) Training via negative sampling.Returns the next representable value ofx1in the direction ofx2, element-wise.<T extends TType>
NextIteration<T> Ops.nextIteration(Operand<T> data) Makes its input available to the next iteration.RandomOps.nonDeterministicInts(Operand<? extends TType> shape) Non-deterministically generates some integers.<U extends TType>
NonDeterministicInts<U> RandomOps.nonDeterministicInts(Operand<? extends TType> shape, Class<U> dtype) Non-deterministically generates some integers.<T extends TNumber>
NonMaxSuppression<T> ImageOps.nonMaxSuppression(Operand<T> boxes, Operand<T> scores, Operand<TInt32> maxOutputSize, Operand<T> iouThreshold, Operand<T> scoreThreshold, Operand<T> softNmsSigma, NonMaxSuppression.Options... options) Greedily selects a subset of bounding boxes in descending order of score, pruning away boxes that have high intersection-over-union (IOU) overlap with previously selected boxes.ImageOps.nonMaxSuppressionWithOverlaps(Operand<TFloat32> overlaps, Operand<TFloat32> scores, Operand<TInt32> maxOutputSize, Operand<TFloat32> overlapThreshold, Operand<TFloat32> scoreThreshold) Greedily selects a subset of bounding boxes in descending order of score, pruning away boxes that have high overlaps with previously selected boxes.DataExperimentalOps.nonSerializableDataset(Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) The ExperimentalNonSerializableDataset operationDataOps.nonSerializableDataset(Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) The NonSerializableDataset operationMathOps.notEqual(Operand<T> x, Operand<T> y, NotEqual.Options... options) Returns the truth value of (x !<T extends TNumber>
NthElement<T> NnOps.nthElement(Operand<T> input, Operand<TInt32> n, NthElement.Options... options) Finds values of then-th order statistic for the last dimension.Ops.oneHot(Operand<? extends TNumber> indices, Operand<TInt32> depth, Operand<U> onValue, Operand<U> offValue, OneHot.Options... options) Returns a one-hot tensor.Creates a one valued tensor given its type and shape.Returns a tensor of ones with the same shape and type as x.DataOps.optimizeDataset(Operand<? extends TType> inputDataset, Operand<TString> optimizationsEnabled, Operand<TString> optimizationsDisabled, Operand<TString> optimizationsDefault, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, OptimizeDataset.Options... options) Creates a dataset by applying related optimizations toinput_dataset.DataOps.optionalGetValue(Operand<? extends TType> optional, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Returns the value stored in an Optional variant or raises an error if none exists.DataOps.optionalHasValue(Operand<? extends TType> optional) Returns true if and only if the given Optional variant has a value.DataOps.optionsDataset(Operand<? extends TType> inputDataset, String serializedOptions, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, OptionsDataset.Options... options) Creates a dataset by attaching tf.data.Options toinput_dataset.Ops.orderedMapPeek(Operand<TInt64> key, Operand<TInt32> indices, List<Class<? extends TType>> dtypes, OrderedMapPeek.Options... options) Op peeks at the values at the specified key.Ops.orderedMapStage(Operand<TInt64> key, Operand<TInt32> indices, Iterable<Operand<?>> values, List<Class<? extends TType>> dtypes, OrderedMapStage.Options... options) Stage (key, values) in the underlying container which behaves like a ordered associative container.Ops.orderedMapUnstage(Operand<TInt64> key, Operand<TInt32> indices, List<Class<? extends TType>> dtypes, OrderedMapUnstage.Options... options) Op removes and returns the values associated with the key from the underlying container.Ops.orderedMapUnstageNoKey(Operand<TInt32> indices, List<Class<? extends TType>> dtypes, OrderedMapUnstageNoKey.Options... options) Op removes and returns the (key, value) element with the smallest key from the underlying container.TpuOps.outfeedDequeueTupleV2(Operand<TInt32> deviceOrdinal, List<Class<? extends TType>> dtypes, List<Shape> shapes) Retrieve multiple values from the computation outfeed.<T extends TType>
OutfeedDequeueV2<T> TpuOps.outfeedDequeueV2(Operand<TInt32> deviceOrdinal, Class<T> dtype, Shape shape) Retrieves a single tensor from the computation outfeed.TpuOps.outfeedEnqueue(Operand<? extends TType> input) Enqueue a Tensor on the computation outfeed.Pads a tensor.DataOps.paddedBatchDataset(Operand<? extends TType> inputDataset, Operand<TInt64> batchSize, Iterable<Operand<TInt64>> paddedShapes, Iterable<Operand<?>> paddingValues, Operand<TBool> dropRemainder, List<Shape> outputShapes, PaddedBatchDataset.Options... options) Creates a dataset that batches and padsbatch_sizeelements from the input.DataOps.parallelBatchDataset(Operand<? extends TType> inputDataset, Operand<TInt64> batchSize, Operand<TInt64> numParallelCalls, Operand<TBool> dropRemainder, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ParallelBatchDataset.Options... options) The ParallelBatchDataset operationDataOps.parallelFilterDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> numParallelCalls, ConcreteFunction predicate, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ParallelFilterDataset.Options... options) Creates a dataset containing elements ofinput_datasetmatchingpredicate.DataExperimentalOps.parallelInterleaveDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> cycleLength, Operand<TInt64> blockLength, Operand<TBool> sloppy, Operand<TInt64> bufferOutputElements, Operand<TInt64> prefetchInputElements, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that appliesfto the outputs ofinput_dataset.DataOps.parallelInterleaveDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> cycleLength, Operand<TInt64> blockLength, Operand<TInt64> bufferOutputElements, Operand<TInt64> prefetchInputElements, Operand<TInt64> numParallelCalls, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ParallelInterleaveDataset.Options... options) Creates a dataset that appliesfto the outputs ofinput_dataset.DataOps.parallelMapDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> numParallelCalls, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ParallelMapDataset.Options... options) Creates a dataset that appliesfto the outputs ofinput_dataset.<U extends TNumber>
ParameterizedTruncatedNormal<U> RandomOps.parameterizedTruncatedNormal(Operand<? extends TNumber> shape, Operand<U> means, Operand<U> stdevs, Operand<U> minvals, Operand<U> maxvals, ParameterizedTruncatedNormal.Options... options) Outputs random values from a normal distribution.IoOps.parseExample(Operand<TString> serialized, Operand<TString> names, Operand<TString> sparseKeys, Operand<TString> denseKeys, Operand<TString> raggedKeys, Iterable<Operand<?>> denseDefaults, Long numSparse, List<Class<? extends TType>> sparseTypes, List<Class<? extends TType>> raggedValueTypes, List<Class<? extends TNumber>> raggedSplitTypes, List<Shape> denseShapes) Transforms a vector of tf.Example protos (as strings) into typed tensors.DataExperimentalOps.parseExampleDataset(Operand<? extends TType> inputDataset, Operand<TInt64> numParallelCalls, Iterable<Operand<?>> denseDefaults, List<String> sparseKeys, List<String> denseKeys, List<Class<? extends TType>> sparseTypes, List<Shape> denseShapes, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ParseExampleDataset.Options... options) Transformsinput_datasetcontainingExampleprotos as vectors of DT_STRING into a dataset ofTensororSparseTensorobjects representing the parsed features.DataOps.parseExampleDataset(Operand<? extends TType> inputDataset, Operand<TInt64> numParallelCalls, Iterable<Operand<?>> denseDefaults, List<String> sparseKeys, List<String> denseKeys, List<Class<? extends TType>> sparseTypes, List<Shape> denseShapes, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, List<Class<? extends TType>> raggedValueTypes, List<Class<? extends TNumber>> raggedSplitTypes, ParseExampleDataset.Options... options) Transformsinput_datasetcontainingExampleprotos as vectors of DT_STRING into a dataset ofTensororSparseTensorobjects representing the parsed features.IoOps.parseSequenceExample(Operand<TString> serialized, Operand<TString> debugName, Operand<TString> contextSparseKeys, Operand<TString> contextDenseKeys, Operand<TString> contextRaggedKeys, Operand<TString> featureListSparseKeys, Operand<TString> featureListDenseKeys, Operand<TString> featureListRaggedKeys, Operand<TBool> featureListDenseMissingAssumedEmpty, Iterable<Operand<?>> contextDenseDefaults, List<Class<? extends TType>> contextSparseTypes, List<Class<? extends TType>> contextRaggedValueTypes, List<Class<? extends TNumber>> contextRaggedSplitTypes, List<Class<? extends TType>> featureListDenseTypes, List<Class<? extends TType>> featureListSparseTypes, List<Class<? extends TType>> featureListRaggedValueTypes, List<Class<? extends TNumber>> featureListRaggedSplitTypes, ParseSequenceExample.Options... options) Transforms a vector of tf.io.SequenceExample protos (as strings) into typed tensors.IoOps.parseSingleExample(Operand<TString> serialized, Iterable<Operand<?>> denseDefaults, Long numSparse, List<String> sparseKeys, List<String> denseKeys, List<Class<? extends TType>> sparseTypes, List<Shape> denseShapes) Transforms a tf.Example proto (as a string) into typed tensors.IoOps.parseSingleSequenceExample(Operand<TString> serialized, Operand<TString> featureListDenseMissingAssumedEmpty, Iterable<Operand<TString>> contextSparseKeys, Iterable<Operand<TString>> contextDenseKeys, Iterable<Operand<TString>> featureListSparseKeys, Iterable<Operand<TString>> featureListDenseKeys, Iterable<Operand<?>> contextDenseDefaults, Operand<TString> debugName, List<Class<? extends TType>> contextSparseTypes, List<Class<? extends TType>> featureListDenseTypes, List<Class<? extends TType>> featureListSparseTypes, ParseSingleSequenceExample.Options... options) Transforms a scalar brain.SequenceExample proto (as strings) into typed tensors.<T extends TType>
ParseTensor<T> IoOps.parseTensor(Operand<TString> serialized, Class<T> outType) Transforms a serialized tensorflow.TensorProto proto into a Tensor.TpuOps.partitionedCall(Iterable<Operand<?>> args, Operand<TInt32> deviceOrdinal, List<Class<? extends TType>> Tout, ConcreteFunction f, PartitionedCall.Options... options) Calls a function placed on a specified TPU device.<T extends TType>
PartitionedOutput<T> TpuOps.partitionedOutput(Operand<T> inputs, Long numSplits, List<Long> partitionDims) An op that demultiplexes a tensor to be sharded by XLA to a list of partitioned outputs outside the XLA computation.<T extends TType>
PlaceholderWithDefault<T> Ops.placeholderWithDefault(Operand<T> input, Shape shape) A placeholder op that passes throughinputwhen its output is not fed.Compute the polygamma function \(\psi^{(n)}(x)\).MathOps.populationCount(Operand<? extends TNumber> x) Computes element-wise population count (a.k.a. popcount, bitsum, bitcount).Computes the power of one value to another.DataOps.prefetchDataset(Operand<? extends TType> inputDataset, Operand<TInt64> bufferSize, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, PrefetchDataset.Options... options) Creates a dataset that asynchronously prefetches elements frominput_dataset.TpuOps.prelinearize(Operand<? extends TType> input, Prelinearize.Options... options) An op which linearizes one Tensor value to an opaque variant tensor.Creates a 1-dimensional operand that represents a new shape containing the dimensions of an operand representing the shape to prepend, followed by the dimensions of an operand representing a shape.<T extends TType>
PreventGradient<T> TrainOps.preventGradient(Operand<T> input, PreventGradient.Options... options) An identity op that triggers an error if a gradient is requested.Ops.print(Operand<TString> input, Print.Options... options) Prints a string scalar.DataExperimentalOps.privateThreadPoolDataset(Operand<? extends TType> inputDataset, Operand<TInt64> numThreads, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that uses a custom thread pool to computeinput_dataset.DataOps.privateThreadPoolDataset(Operand<? extends TType> inputDataset, Operand<TInt64> numThreads, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that uses a custom thread pool to computeinput_dataset.Ops.prod(Operand<T> input, Operand<? extends TNumber> axis, Prod.Options... options) Computes the product of elements across dimensions of a tensor.LinalgOps.qr(Operand<T> input, Qr.Options... options) Computes the QR decompositions of one or more matrices.QuantizationOps.quantize(Operand<TFloat32> input, Operand<TFloat32> minRange, Operand<TFloat32> maxRange, Class<T> T, Quantize.Options... options) Quantize the 'input' tensor of type float to 'output' tensor of type 'T'.<T extends TNumber>
QuantizeAndDequantize<T> QuantizationOps.quantizeAndDequantize(Operand<T> input, Operand<T> inputMin, Operand<T> inputMax, Operand<TInt32> numBits, QuantizeAndDequantize.Options... options) Quantizes then dequantizes a tensor.<T extends TNumber>
QuantizeAndDequantizeV3<T> QuantizationOps.quantizeAndDequantizeV3(Operand<T> input, Operand<T> inputMin, Operand<T> inputMax, Operand<TInt32> numBits, QuantizeAndDequantizeV3.Options... options) Quantizes then dequantizes a tensor.<T extends TNumber>
QuantizeAndDequantizeV4<T> QuantizationOps.quantizeAndDequantizeV4(Operand<T> input, Operand<T> inputMin, Operand<T> inputMax, QuantizeAndDequantizeV4.Options... options) Quantizes then dequantizes a tensor.<T extends TNumber>
QuantizeAndDequantizeV4Grad<T> QuantizationOps.quantizeAndDequantizeV4Grad(Operand<T> gradients, Operand<T> input, Operand<T> inputMin, Operand<T> inputMax, QuantizeAndDequantizeV4Grad.Options... options) Returns the gradient ofQuantizeAndDequantizeV4.<V extends TNumber>
QuantizedAdd<V> MathOps.quantizedAdd(Operand<? extends TNumber> x, Operand<? extends TNumber> y, Operand<TFloat32> minX, Operand<TFloat32> maxX, Operand<TFloat32> minY, Operand<TFloat32> maxY, Class<V> Toutput) Returns x + y element-wise, working on quantized buffers.<T extends TNumber>
QuantizedAvgPool<T> NnOps.quantizedAvgPool(Operand<T> input, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, List<Long> ksize, List<Long> strides, String padding) Produces the average pool of the input tensor for quantized types.<U extends TNumber, T extends TNumber>
QuantizedBatchNormWithGlobalNormalization<U> NnOps.quantizedBatchNormWithGlobalNormalization(Operand<T> t, Operand<TFloat32> tMin, Operand<TFloat32> tMax, Operand<T> m, Operand<TFloat32> mMin, Operand<TFloat32> mMax, Operand<T> v, Operand<TFloat32> vMin, Operand<TFloat32> vMax, Operand<T> beta, Operand<TFloat32> betaMin, Operand<TFloat32> betaMax, Operand<T> gamma, Operand<TFloat32> gammaMin, Operand<TFloat32> gammaMax, Class<U> outType, Float varianceEpsilon, Boolean scaleAfterNormalization) Quantized Batch normalization.<V extends TNumber>
QuantizedBiasAdd<V> NnOps.quantizedBiasAdd(Operand<? extends TNumber> input, Operand<? extends TNumber> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minBias, Operand<TFloat32> maxBias, Class<V> outType) Adds Tensor 'bias' to Tensor 'input' for Quantized types.<T extends TType>
QuantizedConcat<T> QuantizationOps.quantizedConcat(Operand<TInt32> concatDim, Iterable<Operand<T>> values, Iterable<Operand<TFloat32>> inputMins, Iterable<Operand<TFloat32>> inputMaxes) Concatenates quantized tensors along one dimension.<V extends TNumber>
QuantizedConv2d<V> NnOps.quantizedConv2d(Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Class<V> outType, List<Long> strides, String padding, QuantizedConv2d.Options... options) Computes a 2D convolution given quantized 4D input and filter tensors.<V extends TNumber>
QuantizedConv2DAndRelu<V> NnOps.quantizedConv2DAndRelu(Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Class<V> outType, List<Long> strides, String padding, QuantizedConv2DAndRelu.Options... options) The QuantizedConv2DAndRelu operation<V extends TNumber>
QuantizedConv2DAndReluAndRequantize<V> NnOps.quantizedConv2DAndReluAndRequantize(Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Class<V> outType, List<Long> strides, String padding, QuantizedConv2DAndReluAndRequantize.Options... options) The QuantizedConv2DAndReluAndRequantize operation<V extends TNumber>
QuantizedConv2DAndRequantize<V> NnOps.quantizedConv2DAndRequantize(Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Class<V> outType, List<Long> strides, String padding, QuantizedConv2DAndRequantize.Options... options) The QuantizedConv2DAndRequantize operation<V extends TNumber>
QuantizedConv2DPerChannel<V> NnOps.quantizedConv2DPerChannel(Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Class<V> outType, List<Long> strides, String padding, QuantizedConv2DPerChannel.Options... options) Computes QuantizedConv2D per channel.<V extends TNumber>
QuantizedConv2DWithBias<V> NnOps.quantizedConv2DWithBias(Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Class<V> outType, List<Long> strides, String padding, QuantizedConv2DWithBias.Options... options) The QuantizedConv2DWithBias operation<V extends TNumber>
QuantizedConv2DWithBiasAndRelu<V> NnOps.quantizedConv2DWithBiasAndRelu(Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Class<V> outType, List<Long> strides, String padding, QuantizedConv2DWithBiasAndRelu.Options... options) The QuantizedConv2DWithBiasAndRelu operation<W extends TNumber>
QuantizedConv2DWithBiasAndReluAndRequantize<W> NnOps.quantizedConv2DWithBiasAndReluAndRequantize(Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<? extends TNumber> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Class<W> outType, List<Long> strides, String padding, QuantizedConv2DWithBiasAndReluAndRequantize.Options... options) The QuantizedConv2DWithBiasAndReluAndRequantize operation<W extends TNumber>
QuantizedConv2DWithBiasAndRequantize<W> NnOps.quantizedConv2DWithBiasAndRequantize(Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<? extends TNumber> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Class<W> outType, List<Long> strides, String padding, QuantizedConv2DWithBiasAndRequantize.Options... options) The QuantizedConv2DWithBiasAndRequantize operation<X extends TNumber>
QuantizedConv2DWithBiasSignedSumAndReluAndRequantize<X> NnOps.quantizedConv2DWithBiasSignedSumAndReluAndRequantize(Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<? extends TNumber> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Operand<? extends TNumber> summand, Operand<TFloat32> minSummand, Operand<TFloat32> maxSummand, Class<X> outType, List<Long> strides, String padding, QuantizedConv2DWithBiasSignedSumAndReluAndRequantize.Options... options) The QuantizedConv2DWithBiasSignedSumAndReluAndRequantize operation<V extends TNumber>
QuantizedConv2DWithBiasSumAndRelu<V> NnOps.quantizedConv2DWithBiasSumAndRelu(Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Operand<TFloat32> summand, Class<V> outType, List<Long> strides, String padding, QuantizedConv2DWithBiasSumAndRelu.Options... options) The QuantizedConv2DWithBiasSumAndRelu operation<X extends TNumber>
QuantizedConv2DWithBiasSumAndReluAndRequantize<X> NnOps.quantizedConv2DWithBiasSumAndReluAndRequantize(Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<? extends TNumber> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Operand<? extends TNumber> summand, Operand<TFloat32> minSummand, Operand<TFloat32> maxSummand, Class<X> outType, List<Long> strides, String padding, QuantizedConv2DWithBiasSumAndReluAndRequantize.Options... options) The QuantizedConv2DWithBiasSumAndReluAndRequantize operation<V extends TNumber>
QuantizedDepthwiseConv2D<V> NnOps.quantizedDepthwiseConv2D(Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Class<V> outType, List<Long> strides, String padding, QuantizedDepthwiseConv2D.Options... options) Computes quantized depthwise Conv2D.<V extends TNumber>
QuantizedDepthwiseConv2DWithBias<V> NnOps.quantizedDepthwiseConv2DWithBias(Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Class<V> outType, List<Long> strides, String padding, QuantizedDepthwiseConv2DWithBias.Options... options) Computes quantized depthwise Conv2D with Bias.<V extends TNumber>
QuantizedDepthwiseConv2DWithBiasAndRelu<V> NnOps.quantizedDepthwiseConv2DWithBiasAndRelu(Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Class<V> outType, List<Long> strides, String padding, QuantizedDepthwiseConv2DWithBiasAndRelu.Options... options) Computes quantized depthwise Conv2D with Bias and Relu.<W extends TNumber>
QuantizedDepthwiseConv2DWithBiasAndReluAndRequantize<W> NnOps.quantizedDepthwiseConv2DWithBiasAndReluAndRequantize(Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<? extends TNumber> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Class<W> outType, List<Long> strides, String padding, QuantizedDepthwiseConv2DWithBiasAndReluAndRequantize.Options... options) Computes quantized depthwise Conv2D with Bias, Relu and Requantize.<T extends TNumber>
QuantizedInstanceNorm<T> NnOps.quantizedInstanceNorm(Operand<T> x, Operand<TFloat32> xMin, Operand<TFloat32> xMax, QuantizedInstanceNorm.Options... options) Quantized Instance normalization.<V extends TNumber, W extends TNumber>
QuantizedMatMul<V> LinalgOps.quantizedMatMul(Operand<? extends TNumber> a, Operand<? extends TNumber> b, Operand<TFloat32> minA, Operand<TFloat32> maxA, Operand<TFloat32> minB, Operand<TFloat32> maxB, Class<V> Toutput, Class<W> Tactivation, QuantizedMatMul.Options... options) Perform a quantized matrix multiplication ofaby the matrixb.<W extends TNumber>
QuantizedMatMulWithBias<W> LinalgOps.quantizedMatMulWithBias(Operand<? extends TNumber> a, Operand<? extends TNumber> b, Operand<? extends TNumber> bias, Operand<TFloat32> minA, Operand<TFloat32> maxA, Operand<TFloat32> minB, Operand<TFloat32> maxB, Class<W> Toutput, QuantizedMatMulWithBias.Options... options) Performs a quantized matrix multiplication ofaby the matrixbwith bias add.<W extends TNumber>
QuantizedMatMulWithBiasAndDequantize<W> QuantizationOps.quantizedMatMulWithBiasAndDequantize(Operand<? extends TNumber> a, Operand<? extends TNumber> b, Operand<? extends TNumber> bias, Operand<TFloat32> minA, Operand<TFloat32> maxA, Operand<TFloat32> minB, Operand<TFloat32> maxB, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Class<W> Toutput, QuantizedMatMulWithBiasAndDequantize.Options... options) The QuantizedMatMulWithBiasAndDequantize operation<V extends TNumber>
QuantizedMatMulWithBiasAndRelu<V> LinalgOps.quantizedMatMulWithBiasAndRelu(Operand<? extends TNumber> a, Operand<? extends TNumber> b, Operand<TFloat32> bias, Operand<TFloat32> minA, Operand<TFloat32> maxA, Operand<TFloat32> minB, Operand<TFloat32> maxB, Class<V> Toutput, QuantizedMatMulWithBiasAndRelu.Options... options) Perform a quantized matrix multiplication ofaby the matrixbwith bias add and relu fusion.<W extends TNumber>
QuantizedMatMulWithBiasAndReluAndRequantize<W> LinalgOps.quantizedMatMulWithBiasAndReluAndRequantize(Operand<? extends TNumber> a, Operand<? extends TNumber> b, Operand<? extends TNumber> bias, Operand<TFloat32> minA, Operand<TFloat32> maxA, Operand<TFloat32> minB, Operand<TFloat32> maxB, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Class<W> Toutput, QuantizedMatMulWithBiasAndReluAndRequantize.Options... options) Perform a quantized matrix multiplication ofaby the matrixbwith bias add and relu and requantize fusion.<W extends TNumber>
QuantizedMatMulWithBiasAndRequantize<W> QuantizationOps.quantizedMatMulWithBiasAndRequantize(Operand<? extends TNumber> a, Operand<? extends TNumber> b, Operand<? extends TNumber> bias, Operand<TFloat32> minA, Operand<TFloat32> maxA, Operand<TFloat32> minB, Operand<TFloat32> maxB, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Class<W> Toutput, QuantizedMatMulWithBiasAndRequantize.Options... options) The QuantizedMatMulWithBiasAndRequantize operation<T extends TNumber>
QuantizedMaxPool<T> NnOps.quantizedMaxPool(Operand<T> input, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, List<Long> ksize, List<Long> strides, String padding) Produces the max pool of the input tensor for quantized types.<V extends TNumber>
QuantizedMul<V> MathOps.quantizedMul(Operand<? extends TNumber> x, Operand<? extends TNumber> y, Operand<TFloat32> minX, Operand<TFloat32> maxX, Operand<TFloat32> minY, Operand<TFloat32> maxY, Class<V> Toutput) Returns x * y element-wise, working on quantized buffers.<U extends TNumber>
QuantizeDownAndShrinkRange<U> QuantizationOps.quantizeDownAndShrinkRange(Operand<? extends TNumber> input, Operand<TFloat32> inputMin, Operand<TFloat32> inputMax, Class<U> outType) Convert the quantized 'input' tensor into a lower-precision 'output', using the actual distribution of the values to maximize the usage of the lower bit depth and adjusting the output min and max ranges accordingly.<U extends TNumber>
QuantizedRelu<U> NnOps.quantizedRelu(Operand<? extends TNumber> features, Operand<TFloat32> minFeatures, Operand<TFloat32> maxFeatures, Class<U> outType) Computes Quantized Rectified Linear:max(features, 0)<U extends TNumber>
QuantizedRelu6<U> NnOps.quantizedRelu6(Operand<? extends TNumber> features, Operand<TFloat32> minFeatures, Operand<TFloat32> maxFeatures, Class<U> outType) Computes Quantized Rectified Linear 6:min(max(features, 0), 6)<U extends TNumber>
QuantizedReluX<U> NnOps.quantizedReluX(Operand<? extends TNumber> features, Operand<TFloat32> maxValue, Operand<TFloat32> minFeatures, Operand<TFloat32> maxFeatures, Class<U> outType) Computes Quantized Rectified Linear X:min(max(features, 0), max_value)<T extends TType>
QuantizedReshape<T> Ops.quantizedReshape(Operand<T> tensor, Operand<? extends TNumber> shape, Operand<TFloat32> inputMin, Operand<TFloat32> inputMax) Reshapes a quantized tensor as per the Reshape op.<T extends TNumber>
QuantizedResizeBilinear<T> ImageOps.quantizedResizeBilinear(Operand<T> images, Operand<TInt32> sizeOutput, Operand<TFloat32> min, Operand<TFloat32> max, QuantizedResizeBilinear.Options... options) Resize quantizedimagestosizeusing quantized bilinear interpolation.IoOps.queueClose(Operand<? extends TType> handle, QueueClose.Options... options) Closes the given queue.IoOps.queueDequeue(Operand<? extends TType> handle, List<Class<? extends TType>> componentTypes, QueueDequeue.Options... options) Dequeues a tuple of one or more tensors from the given queue.IoOps.queueDequeueMany(Operand<? extends TType> handle, Operand<TInt32> n, List<Class<? extends TType>> componentTypes, QueueDequeueMany.Options... options) Dequeuesntuples of one or more tensors from the given queue.IoOps.queueDequeueUpTo(Operand<? extends TType> handle, Operand<TInt32> n, List<Class<? extends TType>> componentTypes, QueueDequeueUpTo.Options... options) Dequeuesntuples of one or more tensors from the given queue.IoOps.queueEnqueue(Operand<? extends TType> handle, Iterable<Operand<?>> components, QueueEnqueue.Options... options) Enqueues a tuple of one or more tensors in the given queue.IoOps.queueEnqueueMany(Operand<? extends TType> handle, Iterable<Operand<?>> components, QueueEnqueueMany.Options... options) Enqueues zero or more tuples of one or more tensors in the given queue.IoOps.queueIsClosed(Operand<? extends TType> handle) Returns true if queue is closed.Computes the number of elements in the given queue.<U extends TNumber, T extends TNumber>
RaggedBincount<U> RaggedOps.raggedBincount(Operand<TInt64> splits, Operand<T> values, Operand<T> sizeOutput, Operand<U> weights, RaggedBincount.Options... options) Counts the number of occurrences of each value in an integer array.<U extends TNumber>
RaggedCountSparseOutput<U> RaggedOps.raggedCountSparseOutput(Operand<TInt64> splits, Operand<? extends TNumber> values, Operand<U> weights, Boolean binaryOutput, RaggedCountSparseOutput.Options... options) Performs sparse-output bin counting for a ragged tensor input.<T extends TType>
RaggedFillEmptyRows<T> RaggedOps.raggedFillEmptyRows(Operand<TInt64> valueRowids, Operand<T> values, Operand<TInt64> nrows, Operand<T> defaultValue) The RaggedFillEmptyRows operation<T extends TType>
RaggedFillEmptyRowsGrad<T> RaggedOps.raggedFillEmptyRowsGrad(Operand<TInt64> reverseIndexMap, Operand<T> gradValues) The RaggedFillEmptyRowsGrad operation<T extends TNumber, U extends TType>
RaggedGather<T, U> RaggedOps.raggedGather(Iterable<Operand<T>> paramsNestedSplits, Operand<U> paramsDenseValues, Operand<? extends TNumber> indices, Long OUTPUTRAGGEDRANK) Gather ragged slices fromparamsaxis0according toindices.<T extends TNumber>
RaggedRange<TInt64, T> RaggedOps.raggedRange(Operand<T> starts, Operand<T> limits, Operand<T> deltas) Returns aRaggedTensorcontaining the specified sequences of numbers.<U extends TNumber, T extends TNumber>
RaggedRange<U, T> RaggedOps.raggedRange(Operand<T> starts, Operand<T> limits, Operand<T> deltas, Class<U> Tsplits) Returns aRaggedTensorcontaining the specified sequences of numbers.<U extends TType>
RaggedTensorFromVariant<TInt64, U> RaggedOps.raggedTensorFromVariant(Operand<? extends TType> encodedRagged, Long inputRaggedRank, Long outputRaggedRank, Class<U> Tvalues) Decodes avariantTensor into aRaggedTensor.<T extends TNumber, U extends TType>
RaggedTensorFromVariant<T, U> RaggedOps.raggedTensorFromVariant(Operand<? extends TType> encodedRagged, Long inputRaggedRank, Long outputRaggedRank, Class<U> Tvalues, Class<T> Tsplits) Decodes avariantTensor into aRaggedTensor.<U extends TType>
RaggedTensorToSparse<U> RaggedOps.raggedTensorToSparse(Iterable<Operand<? extends TNumber>> rtNestedSplits, Operand<U> rtDenseValues) Converts aRaggedTensorinto aSparseTensorwith the same values.<U extends TType>
RaggedTensorToTensor<U> RaggedOps.raggedTensorToTensor(Operand<? extends TNumber> shape, Operand<U> values, Operand<U> defaultValue, Iterable<Operand<? extends TNumber>> rowPartitionTensors, List<String> rowPartitionTypes) Create a dense tensor from a ragged tensor, possibly altering its shape.RaggedOps.raggedTensorToVariant(Iterable<Operand<? extends TNumber>> rtNestedSplits, Operand<? extends TType> rtDenseValues, Boolean batchedInput) Encodes aRaggedTensorinto avariantTensor.<U extends TType>
RaggedTensorToVariantGradient<U> RaggedOps.raggedTensorToVariantGradient(Operand<? extends TType> encodedRaggedGrad, Operand<? extends TNumber> rowSplits, Operand<TInt32> denseValuesShape, Class<U> Tvalues) Helper used to compute the gradient forRaggedTensorToVariant.<T extends TNumber>
RandomCrop<T> ImageOps.randomCrop(Operand<T> image, Operand<TInt64> sizeOutput, RandomCrop.Options... options) Randomly cropimage.DataExperimentalOps.randomDataset(Operand<TInt64> seed, Operand<TInt64> seed2, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a Dataset that returns pseudorandom numbers.DataOps.randomDataset(Operand<TInt64> seed, Operand<TInt64> seed2, Operand<? extends TType> seedGenerator, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, RandomDataset.Options... options) Creates a Dataset that returns pseudorandom numbers.<U extends TNumber>
RandomGamma<U> RandomOps.randomGamma(Operand<? extends TNumber> shape, Operand<U> alpha, RandomGamma.Options... options) Outputs random values from the Gamma distribution(s) described by alpha.<T extends TNumber>
RandomGammaGrad<T> RandomOps.randomGammaGrad(Operand<T> alpha, Operand<T> sample) Computes the derivative of a Gamma random sample w.r.t.<T extends TNumber>
RandomIndexShuffle<T> Ops.randomIndexShuffle(Operand<T> index, Operand<? extends TNumber> seed, Operand<T> maxIndex, RandomIndexShuffle.Options... options) Outputs the position ofvaluein a permutation of [0, ..., max_index].<V extends TNumber>
RandomPoisson<V> RandomOps.randomPoisson(Operand<? extends TNumber> shape, Operand<? extends TNumber> rate, Class<V> dtype, RandomPoisson.Options... options) Outputs random values from the Poisson distribution(s) described by rate.RandomOps.randomPoisson(Operand<? extends TNumber> shape, Operand<? extends TNumber> rate, RandomPoisson.Options[] options) Outputs random values from the Poisson distribution(s) described by rate.<T extends TType>
RandomShuffle<T> RandomOps.randomShuffle(Operand<T> value, RandomShuffle.Options... options) Randomly shuffles a tensor along its first dimension.<U extends TNumber>
RandomStandardNormal<U> RandomOps.randomStandardNormal(Operand<? extends TNumber> shape, Class<U> dtype, RandomStandardNormal.Options... options) Outputs random values from a normal distribution.<U extends TNumber>
RandomUniform<U> RandomOps.randomUniform(Operand<? extends TNumber> shape, Class<U> dtype, RandomUniform.Options... options) Outputs random values from a uniform distribution.<U extends TNumber>
RandomUniformInt<U> RandomOps.randomUniformInt(Operand<? extends TNumber> shape, Operand<U> minval, Operand<U> maxval, RandomUniformInt.Options... options) Outputs random integers from a uniform distribution.Creates a sequence of numbers.DataOps.rangeDataset(Operand<TInt64> start, Operand<TInt64> stop, Operand<TInt64> step, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, RangeDataset.Options... options) Creates a dataset with a range of values.Returns the rank of a tensor.IoOps.readerNumRecordsProduced(Operand<? extends TType> readerHandle) Returns the number of records this Reader has produced.IoOps.readerNumWorkUnitsCompleted(Operand<? extends TType> readerHandle) Returns the number of work units this Reader has finished processing.IoOps.readerRead(Operand<? extends TType> readerHandle, Operand<? extends TType> queueHandle) Returns the next record (key, value pair) produced by a Reader.IoOps.readerReadUpTo(Operand<? extends TType> readerHandle, Operand<? extends TType> queueHandle, Operand<TInt64> numRecords) Returns up tonum_records(key, value) pairs produced by a Reader.IoOps.readerReset(Operand<? extends TType> readerHandle) Restore a Reader to its initial clean state.IoOps.readerRestoreState(Operand<? extends TType> readerHandle, Operand<TString> state) Restore a reader to a previously saved state.IoOps.readerSerializeState(Operand<? extends TType> readerHandle) Produce a string tensor that encodes the state of a Reader.Reads and outputs the entire contents of the input filename.<T extends TType>
ReadVariableOp<T> Ops.readVariableOp(Operand<? extends TType> resource, Class<T> dtype) Reads the value of a variable.<T extends TType>
ReadVariableSplitND<T> XlaOps.readVariableSplitND(Operand<? extends TType> resource, Class<T> T, Long N, List<Long> numSplits, ReadVariableSplitND.Options... options) Splits resource variable input tensor across all dimensions.Returns the real part of a complex number.Returns the real part of a complex number.Returns x / y element-wise for real types.DataExperimentalOps.rebatchDataset(Operand<? extends TType> inputDataset, Operand<TInt64> numReplicas, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, RebatchDataset.Options... options) Creates a dataset that changes the batch size.DataOps.rebatchDatasetV2(Operand<? extends TType> inputDataset, Operand<TInt64> batchSizes, Operand<TBool> dropRemainder, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that changes the batch size.<T extends TType>
Reciprocal<T> MathOps.reciprocal(Operand<T> x) Computes the reciprocal of x element-wise.<T extends TType>
ReciprocalGrad<T> MathOps.reciprocalGrad(Operand<T> y, Operand<T> dy) Computes the gradient for the inverse ofxwrt its input.Computes the "logical and" of elements across dimensions of a tensor.Computes the "logical or" of elements across dimensions of a tensor.DataOps.reduceDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> initialState, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ReduceDataset.Options... options) Reduces the input dataset to a singleton using a reduce function.ShapeOps.reduceDims(Shape<TInt32> shape, Operand<TInt32> axis) Reduces the shape to the specified axis.ShapeOps.reduceDims(Shape<U> shape, Operand<U> axis, Class<U> type) Reduces the shape to the specified axis.ShapeOps.reduceDims(Operand<T> operand, Operand<TInt32> axis) Reshapes the operand by reducing the shape to the specified axis.ShapeOps.reduceDims(Operand<T> operand, Operand<U> axis, Class<U> type) Reshapes the operand by reducing the shape to the specified axis.StringsOps.reduceJoin(Operand<TString> inputs, Operand<TInt32> reductionIndices, ReduceJoin.Options... options) Joins a string Tensor across the given dimensions.Ops.reduceMax(Operand<T> input, Operand<? extends TNumber> axis, ReduceMax.Options... options) Computes the maximum of elements across dimensions of a tensor.Ops.reduceMin(Operand<T> input, Operand<? extends TNumber> axis, ReduceMin.Options... options) Computes the minimum of elements across dimensions of a tensor.<T extends TType>
ReduceProd<T> Ops.reduceProd(Operand<T> input, Operand<? extends TNumber> axis, ReduceProd.Options... options) Computes the product of elements across dimensions of a tensor.Ops.reduceSum(Operand<T> input, Operand<? extends TNumber> axis, ReduceSum.Options... options) Computes the sum of elements across dimensions of a tensor.Ops.refEnter(Operand<T> data, String frameName, RefEnter.Options... options) Creates or finds a child frame, and makesdataavailable to the child frame.Exits the current frame to its parent frame.<T extends TType>
RefIdentity<T> Ops.refIdentity(Operand<T> input) Return the same ref tensor as the input ref tensor.<T extends TType>
RefNextIteration<T> Ops.refNextIteration(Operand<T> data) Makes its input available to the next iteration.Forwards theindexth element ofinputstooutput.Forwards the ref tensordatato the output port determined bypred.StringsOps.regexFullMatch(Operand<TString> input, Operand<TString> pattern) Check if the input matches the regex pattern.StringsOps.regexReplace(Operand<TString> input, Operand<TString> pattern, Operand<TString> rewrite, RegexReplace.Options... options) Replaces matches of thepatternregular expression ininputwith the replacement string provided inrewrite.DataOps.registerDataset(Operand<? extends TType> dataset, Operand<TString> address, Operand<TString> protocol, Long externalStatePolicy, RegisterDataset.Options... options) Registers a dataset with the tf.data service.The Relayout operation<T extends TType>
RelayoutLike<T> Ops.relayoutLike(Operand<T> input, Operand<? extends TType> layoutInput) The RelayoutLike operationComputes rectified linear:max(features, 0).Computes rectified linear 6:min(max(features, 0), 6).Computes rectified linear 6 gradients for a Relu6 operation.Computes rectified linear gradients for a Relu operation.Ops.remoteCall(Operand<TString> target, Iterable<Operand<?>> args, List<Class<? extends TType>> Tout, ConcreteFunction f) Runs functionfon a remote device indicated bytarget.DataOps.repeatDataset(Operand<? extends TType> inputDataset, Operand<TInt64> count, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, RepeatDataset.Options... options) Creates a dataset that emits the outputs ofinput_datasetcounttimes.<T extends TType>
ReplicatedOutput<T> TpuOps.replicatedOutput(Operand<T> input, Long numReplicas) Connects N outputs from an N-way replicated TPU computation.QuantizationOps.requantizationRange(Operand<? extends TNumber> input, Operand<TFloat32> inputMin, Operand<TFloat32> inputMax) Computes a range that covers the actual values present in a quantized tensor.MathOps.requantizationRangePerChannel(Operand<? extends TNumber> input, Operand<TFloat32> inputMin, Operand<TFloat32> inputMax, Float clipValueMax) Computes requantization range per channel.<U extends TNumber>
Requantize<U> QuantizationOps.requantize(Operand<? extends TNumber> input, Operand<TFloat32> inputMin, Operand<TFloat32> inputMax, Operand<TFloat32> requestedOutputMin, Operand<TFloat32> requestedOutputMax, Class<U> outType) Converts the quantizedinputtensor into a lower-precisionoutput.<U extends TNumber>
RequantizePerChannel<U> MathOps.requantizePerChannel(Operand<? extends TNumber> input, Operand<TFloat32> inputMin, Operand<TFloat32> inputMax, Operand<TFloat32> requestedOutputMin, Operand<TFloat32> requestedOutputMax, Class<U> outType) Requantizes input with min and max values known per channel.Reshapes a tensor.ImageOps.resizeArea(Operand<? extends TNumber> images, Operand<TInt32> sizeOutput, ResizeArea.Options... options) Resizeimagestosizeusing area interpolation.ImageOps.resizeBicubic(Operand<? extends TNumber> images, Operand<TInt32> sizeOutput, ResizeBicubic.Options... options) Resizeimagestosizeusing bicubic interpolation.<T extends TNumber>
ResizeBicubicGrad<T> ImageOps.resizeBicubicGrad(Operand<TFloat32> grads, Operand<T> originalImage, ResizeBicubicGrad.Options... options) Computes the gradient of bicubic interpolation.ImageOps.resizeBilinear(Operand<? extends TNumber> images, Operand<TInt32> sizeOutput, ResizeBilinear.Options... options) Resizeimagestosizeusing bilinear interpolation.<T extends TNumber>
ResizeBilinearGrad<T> ImageOps.resizeBilinearGrad(Operand<TFloat32> grads, Operand<T> originalImage, ResizeBilinearGrad.Options... options) Computes the gradient of bilinear interpolation.<T extends TNumber>
ResizeNearestNeighbor<T> ImageOps.resizeNearestNeighbor(Operand<T> images, Operand<TInt32> sizeOutput, ResizeNearestNeighbor.Options... options) Resizeimagestosizeusing nearest neighbor interpolation.<T extends TNumber>
ResizeNearestNeighborGrad<T> ImageOps.resizeNearestNeighborGrad(Operand<T> grads, Operand<TInt32> sizeOutput, ResizeNearestNeighborGrad.Options... options) Computes the gradient of nearest neighbor interpolation.TrainOps.resourceAccumulatorApplyGradient(Operand<? extends TType> handle, Operand<TInt64> localStep, Operand<? extends TType> gradient) Applies a gradient to a given accumulator.TrainOps.resourceAccumulatorNumAccumulated(Operand<? extends TType> handle) Returns the number of gradients aggregated in the given accumulators.TrainOps.resourceAccumulatorSetGlobalStep(Operand<? extends TType> handle, Operand<TInt64> newGlobalStep) Updates the accumulator with a new value for global_step.<T extends TType>
ResourceAccumulatorTakeGradient<T> TrainOps.resourceAccumulatorTakeGradient(Operand<? extends TType> handle, Operand<TInt32> numRequired, Class<T> dtype) Extracts the average gradient in the given ConditionalAccumulator.<T extends TType>
ResourceApplyAdadeltaTrainOps.resourceApplyAdadelta(Operand<? extends TType> var, Operand<? extends TType> accum, Operand<? extends TType> accumUpdate, Operand<T> lr, Operand<T> rho, Operand<T> epsilon, Operand<T> grad, ResourceApplyAdadelta.Options... options) Update '*var' according to the adadelta scheme.<T extends TType>
ResourceApplyAdagradTrainOps.resourceApplyAdagrad(Operand<? extends TType> var, Operand<? extends TType> accum, Operand<T> lr, Operand<T> epsilon, Operand<T> grad, ResourceApplyAdagrad.Options... options) Update '*var' according to the adagrad scheme.<T extends TType>
ResourceApplyAdagradDaTrainOps.resourceApplyAdagradDa(Operand<? extends TType> var, Operand<? extends TType> gradientAccumulator, Operand<? extends TType> gradientSquaredAccumulator, Operand<T> grad, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<TInt64> globalStep, ResourceApplyAdagradDa.Options... options) Update '*var' according to the proximal adagrad scheme.<T extends TType>
ResourceApplyAdamTrainOps.resourceApplyAdam(Operand<? extends TType> var, Operand<? extends TType> m, Operand<? extends TType> v, Operand<T> beta1Power, Operand<T> beta2Power, Operand<T> lr, Operand<T> beta1, Operand<T> beta2, Operand<T> epsilon, Operand<T> grad, ResourceApplyAdam.Options... options) Update '*var' according to the Adam algorithm.<T extends TType>
ResourceApplyAdaMaxTrainOps.resourceApplyAdaMax(Operand<? extends TType> var, Operand<? extends TType> m, Operand<? extends TType> v, Operand<T> beta1Power, Operand<T> lr, Operand<T> beta1, Operand<T> beta2, Operand<T> epsilon, Operand<T> grad, ResourceApplyAdaMax.Options... options) Update '*var' according to the AdaMax algorithm.<T extends TType>
ResourceApplyAdamWithAmsgradTrainOps.resourceApplyAdamWithAmsgrad(Operand<? extends TType> var, Operand<? extends TType> m, Operand<? extends TType> v, Operand<? extends TType> vhat, Operand<T> beta1Power, Operand<T> beta2Power, Operand<T> lr, Operand<T> beta1, Operand<T> beta2, Operand<T> epsilon, Operand<T> grad, ResourceApplyAdamWithAmsgrad.Options... options) Update '*var' according to the Adam algorithm.<T extends TType>
ResourceApplyAddSignTrainOps.resourceApplyAddSign(Operand<? extends TType> var, Operand<? extends TType> m, Operand<T> lr, Operand<T> alpha, Operand<T> signDecay, Operand<T> beta, Operand<T> grad, ResourceApplyAddSign.Options... options) Update '*var' according to the AddSign update.<T extends TType>
ResourceApplyCenteredRmsPropTrainOps.resourceApplyCenteredRmsProp(Operand<? extends TType> var, Operand<? extends TType> mg, Operand<? extends TType> ms, Operand<? extends TType> mom, Operand<T> lr, Operand<T> rho, Operand<T> momentum, Operand<T> epsilon, Operand<T> grad, ResourceApplyCenteredRmsProp.Options... options) Update '*var' according to the centered RMSProp algorithm.<T extends TType>
ResourceApplyFtrlTrainOps.resourceApplyFtrl(Operand<? extends TType> var, Operand<? extends TType> accum, Operand<? extends TType> linear, Operand<T> grad, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<T> l2Shrinkage, Operand<T> lrPower, ResourceApplyFtrl.Options... options) Update '*var' according to the Ftrl-proximal scheme.<T extends TType>
ResourceApplyGradientDescentTrainOps.resourceApplyGradientDescent(Operand<? extends TType> var, Operand<T> alpha, Operand<T> delta, ResourceApplyGradientDescent.Options... options) Update '*var' by subtracting 'alpha' * 'delta' from it.<T extends TType>
ResourceApplyKerasMomentumTrainOps.resourceApplyKerasMomentum(Operand<? extends TType> var, Operand<? extends TType> accum, Operand<T> lr, Operand<T> grad, Operand<T> momentum, ResourceApplyKerasMomentum.Options... options) Update '*var' according to the momentum scheme.<T extends TType>
ResourceApplyMomentumTrainOps.resourceApplyMomentum(Operand<? extends TType> var, Operand<? extends TType> accum, Operand<T> lr, Operand<T> grad, Operand<T> momentum, ResourceApplyMomentum.Options... options) Update '*var' according to the momentum scheme.<T extends TType>
ResourceApplyPowerSignTrainOps.resourceApplyPowerSign(Operand<? extends TType> var, Operand<? extends TType> m, Operand<T> lr, Operand<T> logbase, Operand<T> signDecay, Operand<T> beta, Operand<T> grad, ResourceApplyPowerSign.Options... options) Update '*var' according to the AddSign update.<T extends TType>
ResourceApplyProximalAdagradTrainOps.resourceApplyProximalAdagrad(Operand<? extends TType> var, Operand<? extends TType> accum, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<T> grad, ResourceApplyProximalAdagrad.Options... options) Update '*var' and '*accum' according to FOBOS with Adagrad learning rate.<T extends TType>
ResourceApplyProximalGradientDescentTrainOps.resourceApplyProximalGradientDescent(Operand<? extends TType> var, Operand<T> alpha, Operand<T> l1, Operand<T> l2, Operand<T> delta, ResourceApplyProximalGradientDescent.Options... options) Update '*var' as FOBOS algorithm with fixed learning rate.<T extends TType>
ResourceApplyRmsPropTrainOps.resourceApplyRmsProp(Operand<? extends TType> var, Operand<? extends TType> ms, Operand<? extends TType> mom, Operand<T> lr, Operand<T> rho, Operand<T> momentum, Operand<T> epsilon, Operand<T> grad, ResourceApplyRmsProp.Options... options) Update '*var' according to the RMSProp algorithm.<T extends TNumber>
ResourceCountUpTo<T> Ops.resourceCountUpTo(Operand<? extends TType> resource, Long limit, Class<T> T) Increments variable pointed to by 'resource' until it reaches 'limit'.<U extends TType>
ResourceGather<U> Ops.resourceGather(Operand<? extends TType> resource, Operand<? extends TNumber> indices, Class<U> dtype, ResourceGather.Options... options) Gather slices from the variable pointed to byresourceaccording toindices.<U extends TType>
ResourceGatherNd<U> Ops.resourceGatherNd(Operand<? extends TType> resource, Operand<? extends TNumber> indices, Class<U> dtype) The ResourceGatherNd operationOps.resourceScatterAdd(Operand<? extends TType> resource, Operand<? extends TNumber> indices, Operand<? extends TType> updates) Adds sparse updates to the variable referenced byresource.Ops.resourceScatterDiv(Operand<? extends TType> resource, Operand<? extends TNumber> indices, Operand<? extends TType> updates) Divides sparse updates into the variable referenced byresource.Ops.resourceScatterMax(Operand<? extends TType> resource, Operand<? extends TNumber> indices, Operand<? extends TType> updates) Reduces sparse updates into the variable referenced byresourceusing themaxoperation.Ops.resourceScatterMin(Operand<? extends TType> resource, Operand<? extends TNumber> indices, Operand<? extends TType> updates) Reduces sparse updates into the variable referenced byresourceusing theminoperation.Ops.resourceScatterMul(Operand<? extends TType> resource, Operand<? extends TNumber> indices, Operand<? extends TType> updates) Multiplies sparse updates into the variable referenced byresource.Ops.resourceScatterNdAdd(Operand<? extends TType> ref, Operand<? extends TNumber> indices, Operand<? extends TType> updates, ResourceScatterNdAdd.Options... options) Applies sparse addition to individual values or slices in a Variable.Ops.resourceScatterNdMax(Operand<? extends TType> ref, Operand<? extends TNumber> indices, Operand<? extends TType> updates, ResourceScatterNdMax.Options... options) The ResourceScatterNdMax operationOps.resourceScatterNdMin(Operand<? extends TType> ref, Operand<? extends TNumber> indices, Operand<? extends TType> updates, ResourceScatterNdMin.Options... options) The ResourceScatterNdMin operationOps.resourceScatterNdSub(Operand<? extends TType> ref, Operand<? extends TNumber> indices, Operand<? extends TType> updates, ResourceScatterNdSub.Options... options) Applies sparse subtraction to individual values or slices in a Variable.Ops.resourceScatterNdUpdate(Operand<? extends TType> ref, Operand<? extends TNumber> indices, Operand<? extends TType> updates, ResourceScatterNdUpdate.Options... options) Applies sparseupdatesto individual values or slices within a given variable according toindices.Ops.resourceScatterSub(Operand<? extends TType> resource, Operand<? extends TNumber> indices, Operand<? extends TType> updates) Subtracts sparse updates from the variable referenced byresource.Ops.resourceScatterUpdate(Operand<? extends TType> resource, Operand<? extends TNumber> indices, Operand<? extends TType> updates) Assigns sparse updates to the variable referenced byresource.<T extends TType>
ResourceSparseApplyAdadeltaTrainOps.resourceSparseApplyAdadelta(Operand<? extends TType> var, Operand<? extends TType> accum, Operand<? extends TType> accumUpdate, Operand<T> lr, Operand<T> rho, Operand<T> epsilon, Operand<T> grad, Operand<? extends TNumber> indices, ResourceSparseApplyAdadelta.Options... options) var: Should be from a Variable().<T extends TType>
ResourceSparseApplyAdagradTrainOps.resourceSparseApplyAdagrad(Operand<? extends TType> var, Operand<? extends TType> accum, Operand<T> lr, Operand<T> grad, Operand<? extends TNumber> indices, ResourceSparseApplyAdagrad.Options... options) Update relevant entries in '*var' and '*accum' according to the adagrad scheme.<T extends TType>
ResourceSparseApplyAdagradDaTrainOps.resourceSparseApplyAdagradDa(Operand<? extends TType> var, Operand<? extends TType> gradientAccumulator, Operand<? extends TType> gradientSquaredAccumulator, Operand<T> grad, Operand<? extends TNumber> indices, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<TInt64> globalStep, ResourceSparseApplyAdagradDa.Options... options) Update entries in '*var' and '*accum' according to the proximal adagrad scheme.<T extends TType>
ResourceSparseApplyAdagradV2TrainOps.resourceSparseApplyAdagradV2(Operand<? extends TType> var, Operand<? extends TType> accum, Operand<T> lr, Operand<T> epsilon, Operand<T> grad, Operand<? extends TNumber> indices, ResourceSparseApplyAdagradV2.Options... options) Update relevant entries in '*var' and '*accum' according to the adagrad scheme.<T extends TType>
ResourceSparseApplyCenteredRmsPropTrainOps.resourceSparseApplyCenteredRmsProp(Operand<? extends TType> var, Operand<? extends TType> mg, Operand<? extends TType> ms, Operand<? extends TType> mom, Operand<T> lr, Operand<T> rho, Operand<T> momentum, Operand<T> epsilon, Operand<T> grad, Operand<? extends TNumber> indices, ResourceSparseApplyCenteredRmsProp.Options... options) Update '*var' according to the centered RMSProp algorithm.<T extends TType>
ResourceSparseApplyFtrlTrainOps.resourceSparseApplyFtrl(Operand<? extends TType> var, Operand<? extends TType> accum, Operand<? extends TType> linear, Operand<T> grad, Operand<? extends TNumber> indices, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<T> l2Shrinkage, Operand<T> lrPower, ResourceSparseApplyFtrl.Options... options) Update relevant entries in '*var' according to the Ftrl-proximal scheme.<T extends TType>
ResourceSparseApplyKerasMomentumTrainOps.resourceSparseApplyKerasMomentum(Operand<? extends TType> var, Operand<? extends TType> accum, Operand<T> lr, Operand<T> grad, Operand<? extends TNumber> indices, Operand<T> momentum, ResourceSparseApplyKerasMomentum.Options... options) Update relevant entries in '*var' and '*accum' according to the momentum scheme.<T extends TType>
ResourceSparseApplyMomentumTrainOps.resourceSparseApplyMomentum(Operand<? extends TType> var, Operand<? extends TType> accum, Operand<T> lr, Operand<T> grad, Operand<? extends TNumber> indices, Operand<T> momentum, ResourceSparseApplyMomentum.Options... options) Update relevant entries in '*var' and '*accum' according to the momentum scheme.<T extends TType>
ResourceSparseApplyProximalAdagradTrainOps.resourceSparseApplyProximalAdagrad(Operand<? extends TType> var, Operand<? extends TType> accum, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<T> grad, Operand<? extends TNumber> indices, ResourceSparseApplyProximalAdagrad.Options... options) Sparse update entries in '*var' and '*accum' according to FOBOS algorithm.<T extends TType>
ResourceSparseApplyProximalGradientDescentTrainOps.resourceSparseApplyProximalGradientDescent(Operand<? extends TType> var, Operand<T> alpha, Operand<T> l1, Operand<T> l2, Operand<T> grad, Operand<? extends TNumber> indices, ResourceSparseApplyProximalGradientDescent.Options... options) Sparse update '*var' as FOBOS algorithm with fixed learning rate.<T extends TType>
ResourceSparseApplyRmsPropTrainOps.resourceSparseApplyRmsProp(Operand<? extends TType> var, Operand<? extends TType> ms, Operand<? extends TType> mom, Operand<T> lr, Operand<T> rho, Operand<T> momentum, Operand<T> epsilon, Operand<T> grad, Operand<? extends TNumber> indices, ResourceSparseApplyRmsProp.Options... options) Update '*var' according to the RMSProp algorithm.<T extends TNumber>
ResourceStridedSliceAssignOps.resourceStridedSliceAssign(Operand<? extends TType> ref, Operand<T> begin, Operand<T> end, Operand<T> strides, Operand<? extends TType> value, ResourceStridedSliceAssign.Options... options) Assignvalueto the sliced l-value reference ofref.TrainOps.restore(Operand<TString> prefix, Operand<TString> tensorNames, Operand<TString> shapeAndSlices, List<Class<? extends TType>> dtypes) Restores tensors from a V2 checkpoint.<T extends TType>
RestoreSlice<T> TrainOps.restoreSlice(Operand<TString> filePattern, Operand<TString> tensorName, Operand<TString> shapeAndSlice, Class<T> dt, RestoreSlice.Options... options) Restores a tensor from checkpoint files.Reverses specific dimensions of a tensor.<T extends TType>
ReverseSequence<T> Ops.reverseSequence(Operand<T> input, Operand<? extends TNumber> seqLengths, Long seqDim, ReverseSequence.Options... options) Reverses variable length slices.DataOps.rewriteDataset(Operand<? extends TType> inputDataset, Operand<TString> rewriteName, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) The RewriteDataset operationReal-valued fast Fourier transform.2D real-valued fast Fourier transform.3D real-valued fast Fourier transform.SignalOps.rfftNd(Operand<? extends TNumber> input, Operand<TInt32> fftLength, Operand<TInt32> axes, Class<U> Tcomplex) ND fast real Fourier transform.Converts one or more images from RGB to HSV.<T extends TNumber>
RightShift<T> BitwiseOps.rightShift(Operand<T> x, Operand<T> y) Elementwise computes the bitwise right-shift ofxandy.Returns element-wise integer closest to x.RandomOps.rngReadAndSkip(Operand<? extends TType> resource, Operand<TInt32> alg, Operand<? extends TType> delta) Advance the counter of a counter-based RNG.RandomOps.rngSkip(Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<TInt64> delta) Advance the counter of a counter-based RNG.Rolls the elements of a tensor along an axis.Rounds the values of a tensor to the nearest integer, element-wise.Computes reciprocal of square root of x element-wise.Computes the gradient for the rsqrt ofxwrt its input.<T extends TNumber>
SampleDistortedBoundingBox<T> ImageOps.sampleDistortedBoundingBox(Operand<T> imageSize, Operand<TFloat32> boundingBoxes, Operand<TFloat32> minObjectCovered, SampleDistortedBoundingBox.Options... options) Generate a single randomly distorted bounding box for an image.DataOps.samplingDataset(Operand<? extends TType> inputDataset, Operand<TFloat32> rate, Operand<TInt64> seed, Operand<TInt64> seed2, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that takes a Bernoulli sample of the contents of another dataset.TrainOps.save(Operand<TString> prefix, Operand<TString> tensorNames, Operand<TString> shapeAndSlices, Iterable<Operand<?>> tensors) Saves tensors in V2 checkpoint format.DataOps.saveDataset(Operand<? extends TType> inputDataset, Operand<TString> path, Iterable<Operand<?>> shardFuncOtherArgs, ConcreteFunction shardFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, SaveDataset.Options... options) The SaveDatasetV2 operationTrainOps.saveSlices(Operand<TString> filename, Operand<TString> tensorNames, Operand<TString> shapesAndSlices, Iterable<Operand<?>> data) Saves input tensors slices to disk.SummaryOps.scalarSummary(Operand<TString> tags, Operand<? extends TNumber> values) Outputs aSummaryprotocol buffer with scalar values.ImageOps.scaleAndTranslate(Operand<? extends TNumber> images, Operand<TInt32> sizeOutput, Operand<TFloat32> scale, Operand<TFloat32> translation, ScaleAndTranslate.Options... options) The ScaleAndTranslate operation<T extends TNumber>
ScaleAndTranslateGrad<T> ImageOps.scaleAndTranslateGrad(Operand<T> grads, Operand<T> originalImage, Operand<TFloat32> scale, Operand<TFloat32> translation, ScaleAndTranslateGrad.Options... options) The ScaleAndTranslateGrad operationDataExperimentalOps.scanDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> initialState, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ScanDataset.Options... options) Creates a dataset successively reducesfover the elements ofinput_dataset.DataOps.scanDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> initialState, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ScanDataset.Options... options) Creates a dataset successively reducesfover the elements ofinput_dataset.<T extends TType>
ScatterAdd<T> Ops.scatterAdd(Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterAdd.Options... options) Adds sparse updates to a variable reference.<T extends TType>
ScatterDiv<T> Ops.scatterDiv(Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterDiv.Options... options) Divides a variable reference by sparse updates.<T extends TNumber>
ScatterMax<T> Ops.scatterMax(Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterMax.Options... options) Reduces sparse updates into a variable reference using themaxoperation.<T extends TNumber>
ScatterMin<T> Ops.scatterMin(Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterMin.Options... options) Reduces sparse updates into a variable reference using theminoperation.<T extends TType>
ScatterMul<T> Ops.scatterMul(Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterMul.Options... options) Multiplies sparse updates into a variable reference.Ops.scatterNd(Operand<T> indices, Operand<U> updates, Operand<T> shape, ScatterNd.Options... options) Scattersupdatesinto a tensor of shapeshapeaccording toindices.<T extends TType>
ScatterNdAdd<T> Ops.scatterNdAdd(Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterNdAdd.Options... options) Applies sparse addition to individual values or slices in a Variable.<T extends TType>
ScatterNdMax<T> Ops.scatterNdMax(Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterNdMax.Options... options) Computes element-wise maximum.<T extends TType>
ScatterNdMin<T> Ops.scatterNdMin(Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterNdMin.Options... options) Computes element-wise minimum.<T extends TType>
ScatterNdNonAliasingAdd<T> Ops.scatterNdNonAliasingAdd(Operand<T> input, Operand<? extends TNumber> indices, Operand<T> updates, ScatterNdNonAliasingAdd.Options... options) Applies sparse addition toinputusing individual values or slices fromupdatesaccording to indicesindices.<T extends TType>
ScatterNdSub<T> Ops.scatterNdSub(Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterNdSub.Options... options) Applies sparse subtraction to individual values or slices in a Variable.<T extends TType>
ScatterNdUpdate<T> Ops.scatterNdUpdate(Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterNdUpdate.Options... options) Applies sparseupdatesto individual values or slices within a given variable according toindices.<T extends TType>
ScatterSub<T> Ops.scatterSub(Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterSub.Options... options) Subtracts sparse updates to a variable reference.<T extends TType>
ScatterUpdate<T> Ops.scatterUpdate(Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterUpdate.Options... options) Applies sparse updates to a variable reference.TrainOps.sdcaFprint(Operand<TString> input) Computes fingerprints of the input strings.TrainOps.sdcaOptimizer(Iterable<Operand<TInt64>> sparseExampleIndices, Iterable<Operand<TInt64>> sparseFeatureIndices, Iterable<Operand<TFloat32>> sparseFeatureValues, Iterable<Operand<TFloat32>> denseFeatures, Operand<TFloat32> exampleWeights, Operand<TFloat32> exampleLabels, Iterable<Operand<TInt64>> sparseIndices, Iterable<Operand<TFloat32>> sparseWeights, Iterable<Operand<TFloat32>> denseWeights, Operand<TFloat32> exampleStateData, String lossType, Float l1, Float l2, Long numLossPartitions, Long numInnerIterations, SdcaOptimizer.Options... options) Distributed version of Stochastic Dual Coordinate Ascent (SDCA) optimizer for linear models with L1 + L2 regularization.<T extends TNumber>
SegmentMax<T> MathOps.segmentMax(Operand<T> data, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments) Computes the maximum along segments of a tensor.<T extends TType>
SegmentMean<T> MathOps.segmentMean(Operand<T> data, Operand<? extends TNumber> segmentIds) Computes the mean along segments of a tensor.<T extends TNumber>
SegmentMin<T> MathOps.segmentMin(Operand<T> data, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments) Computes the minimum along segments of a tensor.<T extends TType>
SegmentProd<T> MathOps.segmentProd(Operand<T> data, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments) Computes the product along segments of a tensor.<T extends TType>
SegmentSum<T> MathOps.segmentSum(Operand<T> data, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments) Computes the sum along segments of a tensor.The SelectV2 operation<T extends TType>
SelfAdjointEig<T> LinalgOps.selfAdjointEig(Operand<T> input, SelfAdjointEig.Options... options) Computes the eigen decomposition of one or more square self-adjoint matrices.Computes scaled exponential linear:scale * alpha * (exp(features) - 1)if < 0,scale * featuresotherwise.Computes gradients for the scaled exponential linear (Selu) operation.Ops.send(Operand<? extends TType> tensor, String tensorName, String sendDevice, Long sendDeviceIncarnation, String recvDevice, Send.Options... options) Sends the named tensor from send_device to recv_device.DataOps.serializeIterator(Operand<? extends TType> resourceHandle, SerializeIterator.Options... options) Converts the givenresource_handlerepresenting an iterator to a variant tensor.IoOps.serializeManySparse(Operand<TInt64> sparseIndices, Operand<? extends TType> sparseValues, Operand<TInt64> sparseShape) Serialize anN-minibatchSparseTensorinto an[N, 3]Tensorobject.<U extends TType>
SerializeManySparse<U> IoOps.serializeManySparse(Operand<TInt64> sparseIndices, Operand<? extends TType> sparseValues, Operand<TInt64> sparseShape, Class<U> outType) Serialize anN-minibatchSparseTensorinto an[N, 3]Tensorobject.IoOps.serializeSparse(Operand<TInt64> sparseIndices, Operand<? extends TType> sparseValues, Operand<TInt64> sparseShape) Serialize aSparseTensorinto a[3]Tensorobject.<U extends TType>
SerializeSparse<U> IoOps.serializeSparse(Operand<TInt64> sparseIndices, Operand<? extends TType> sparseValues, Operand<TInt64> sparseShape, Class<U> outType) Serialize aSparseTensorinto a[3]Tensorobject.IoOps.serializeTensor(Operand<? extends TType> tensor) Transforms a Tensor into a serialized TensorProto proto.Computes the difference between two lists of numbers or strings.Computes the difference between two lists of numbers or strings.Ops.setSize(Operand<TInt64> setIndices, Operand<? extends TType> setValues, Operand<TInt64> setShape, SetSize.Options... options) Number of unique elements along last dimension of inputset.DataExperimentalOps.setStatsAggregatorDataset(Operand<? extends TType> inputDataset, Operand<? extends TType> statsAggregator, Operand<TString> tag, Operand<TString> counterPrefix, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) The ExperimentalSetStatsAggregatorDataset operationDataOps.setStatsAggregatorDataset(Operand<? extends TType> inputDataset, Operand<? extends TType> statsAggregator, Operand<TString> tag, Operand<TString> counterPrefix, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) The SetStatsAggregatorDataset operationReturns the shape of a tensor.Returns the shape of a tensor.DataOps.shardDataset(Operand<? extends TType> inputDataset, Operand<TInt64> numShards, Operand<TInt64> index, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ShardDataset.Options... options) Creates aDatasetthat includes only 1/num_shardsof this dataset.Generate a sharded filename.IoOps.shardedFilespec(Operand<TString> basename, Operand<TInt32> numShards) Generate a glob pattern matching all sharded file names.DataOps.shuffleAndRepeatDataset(Operand<? extends TType> inputDataset, Operand<TInt64> bufferSize, Operand<TInt64> seed, Operand<TInt64> seed2, Operand<TInt64> count, Operand<? extends TType> seedGenerator, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ShuffleAndRepeatDataset.Options... options) The ShuffleAndRepeatDatasetV2 operationDataOps.shuffleDataset(Operand<? extends TType> inputDataset, Operand<TInt64> bufferSize, Operand<TInt64> seed, Operand<TInt64> seed2, Operand<? extends TType> seedGenerator, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ShuffleDataset.Options... options) The ShuffleDatasetV3 operationComputes sigmoid ofxelement-wise.<T extends TType>
SigmoidGrad<T> MathOps.sigmoidGrad(Operand<T> y, Operand<T> dy) Computes the gradient of the sigmoid ofxwrt its input.Returns an element-wise indication of the sign of a number.Computes sine of x element-wise.Computes hyperbolic sine of x element-wise.Returns the size of a tensor.Returns the size of a tensor.Get the size of the specified dimension in the shape.Get the size of the specified dimension in the shape.Get the size of the specified dimension for the shape of the tensor.Get the size of the specified dimension for the shape of the tensor.DataOps.skipDataset(Operand<? extends TType> inputDataset, Operand<TInt64> count, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, SkipDataset.Options... options) Creates a dataset that skipscountelements from theinput_dataset.DataExperimentalOps.sleepDataset(Operand<? extends TType> inputDataset, Operand<TInt64> sleepMicroseconds, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) The ExperimentalSleepDataset operationDataOps.sleepDataset(Operand<? extends TType> inputDataset, Operand<TInt64> sleepMicroseconds, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) The SleepDataset operationReturn a slice from 'input'.DataExperimentalOps.slidingWindowDataset(Operand<? extends TType> inputDataset, Operand<TInt64> windowSize, Operand<TInt64> windowShift, Operand<TInt64> windowStride, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that passes a sliding window overinput_dataset.DataOps.slidingWindowDataset(Operand<? extends TType> inputDataset, Operand<TInt64> windowSize, Operand<TInt64> windowShift, Operand<TInt64> windowStride, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, SlidingWindowDataset.Options... options) Creates a dataset that passes a sliding window overinput_dataset.Returns a copy of the input tensor.DataOps.snapshotChunkDataset(Operand<TString> chunkFile, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, SnapshotChunkDataset.Options... options) The SnapshotChunkDataset operationDataOps.snapshotDataset(Operand<? extends TType> inputDataset, Operand<TString> path, Iterable<Operand<?>> readerFuncOtherArgs, Iterable<Operand<?>> shardFuncOtherArgs, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ConcreteFunction readerFunc, ConcreteFunction shardFunc, SnapshotDataset.Options... options) Creates a dataset that will write to / read from a snapshot.DataOps.snapshotDatasetReader(Operand<TString> shardDir, Operand<TInt64> startIndex, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, Long version, SnapshotDatasetReader.Options... options) The SnapshotDatasetReader operationGenerates points from the Sobol sequence.<T extends TNumber>
SobolSample<T> MathOps.sobolSample(Operand<TInt32> dim, Operand<TInt32> numResults, Operand<TInt32> skip, Class<T> dtype) Generates points from the Sobol sequence.Computes softmax activations.<T extends TNumber>
SoftmaxCrossEntropyWithLogits<T> NnOps.softmaxCrossEntropyWithLogits(Operand<T> features, Operand<T> labels) Computes softmax cross entropy cost and gradients to backpropagate.The Softplus operation<T extends TNumber>
SoftplusGrad<T> MathOps.softplusGrad(Operand<T> gradients, Operand<T> features) Computes softplus gradients for a softplus operation.Computes softsign:features / (abs(features) + 1).<T extends TNumber>
SoftsignGrad<T> NnOps.softsignGrad(Operand<T> gradients, Operand<T> features) Computes softsign gradients for a softsign operation.LinalgOps.solve(Operand<T> matrix, Operand<T> rhs, Solve.Options... options) Solves systems of linear equations.<T extends TType>
SpaceToBatch<T> NnOps.spaceToBatch(Operand<T> input, Operand<? extends TNumber> paddings, Long blockSize) SpaceToBatch for 4-D tensors of type T.<T extends TType>
SpaceToBatchNd<T> Ops.spaceToBatchNd(Operand<T> input, Operand<? extends TNumber> blockShape, Operand<? extends TNumber> paddings) SpaceToBatch for N-D tensors of type T.<T extends TType>
SpaceToDepth<T> NnOps.spaceToDepth(Operand<T> input, Long blockSize, SpaceToDepth.Options... options) SpaceToDepth for tensors of type T.SparseOps.sparseAccumulatorApplyGradient(Operand<TString> handle, Operand<TInt64> localStep, Operand<TInt64> gradientIndices, Operand<? extends TType> gradientValues, Operand<TInt64> gradientShape, Boolean hasKnownShape) Applies a sparse gradient to a given accumulator.<T extends TType>
SparseAccumulatorTakeGradient<T> SparseOps.sparseAccumulatorTakeGradient(Operand<TString> handle, Operand<TInt32> numRequired, Class<T> dtype) Extracts the average sparse gradient in a SparseConditionalAccumulator.SparseOps.sparseAdd(Operand<TInt64> aIndices, Operand<T> aValues, Operand<TInt64> aShape, Operand<TInt64> bIndices, Operand<T> bValues, Operand<TInt64> bShape, Operand<? extends TNumber> thresh) Adds twoSparseTensorobjects to produce anotherSparseTensor.<T extends TType>
SparseAddGrad<T> SparseOps.sparseAddGrad(Operand<T> backpropValGrad, Operand<TInt64> aIndices, Operand<TInt64> bIndices, Operand<TInt64> sumIndices) The gradient operator for the SparseAdd op.<T extends TType>
SparseApplyAdadelta<T> TrainOps.sparseApplyAdadelta(Operand<T> var, Operand<T> accum, Operand<T> accumUpdate, Operand<T> lr, Operand<T> rho, Operand<T> epsilon, Operand<T> grad, Operand<? extends TNumber> indices, SparseApplyAdadelta.Options... options) var: Should be from a Variable().<T extends TType>
SparseApplyAdagrad<T> TrainOps.sparseApplyAdagrad(Operand<T> var, Operand<T> accum, Operand<T> lr, Operand<T> epsilon, Operand<T> grad, Operand<? extends TNumber> indices, SparseApplyAdagrad.Options... options) Update relevant entries in '*var' and '*accum' according to the adagrad scheme.<T extends TType>
SparseApplyAdagradDa<T> TrainOps.sparseApplyAdagradDa(Operand<T> var, Operand<T> gradientAccumulator, Operand<T> gradientSquaredAccumulator, Operand<T> grad, Operand<? extends TNumber> indices, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<TInt64> globalStep, SparseApplyAdagradDa.Options... options) Update entries in '*var' and '*accum' according to the proximal adagrad scheme.<T extends TType>
SparseApplyCenteredRmsProp<T> TrainOps.sparseApplyCenteredRmsProp(Operand<T> var, Operand<T> mg, Operand<T> ms, Operand<T> mom, Operand<T> lr, Operand<T> rho, Operand<T> momentum, Operand<T> epsilon, Operand<T> grad, Operand<? extends TNumber> indices, SparseApplyCenteredRmsProp.Options... options) Update '*var' according to the centered RMSProp algorithm.<T extends TType>
SparseApplyFtrl<T> TrainOps.sparseApplyFtrl(Operand<T> var, Operand<T> accum, Operand<T> linear, Operand<T> grad, Operand<? extends TNumber> indices, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<T> l2Shrinkage, Operand<T> lrPower, SparseApplyFtrl.Options... options) Update relevant entries in '*var' according to the Ftrl-proximal scheme.<T extends TType>
SparseApplyMomentum<T> TrainOps.sparseApplyMomentum(Operand<T> var, Operand<T> accum, Operand<T> lr, Operand<T> grad, Operand<? extends TNumber> indices, Operand<T> momentum, SparseApplyMomentum.Options... options) Update relevant entries in '*var' and '*accum' according to the momentum scheme.<T extends TType>
SparseApplyProximalAdagrad<T> TrainOps.sparseApplyProximalAdagrad(Operand<T> var, Operand<T> accum, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<T> grad, Operand<? extends TNumber> indices, SparseApplyProximalAdagrad.Options... options) Sparse update entries in '*var' and '*accum' according to FOBOS algorithm.<T extends TType>
SparseApplyProximalGradientDescent<T> TrainOps.sparseApplyProximalGradientDescent(Operand<T> var, Operand<T> alpha, Operand<T> l1, Operand<T> l2, Operand<T> grad, Operand<? extends TNumber> indices, SparseApplyProximalGradientDescent.Options... options) Sparse update '*var' as FOBOS algorithm with fixed learning rate.<T extends TType>
SparseApplyRmsProp<T> TrainOps.sparseApplyRmsProp(Operand<T> var, Operand<T> ms, Operand<T> mom, Operand<T> lr, Operand<T> rho, Operand<T> momentum, Operand<T> epsilon, Operand<T> grad, Operand<? extends TNumber> indices, SparseApplyRmsProp.Options... options) Update '*var' according to the RMSProp algorithm.<U extends TNumber, T extends TNumber>
SparseBincount<U> SparseOps.sparseBincount(Operand<TInt64> indices, Operand<T> values, Operand<TInt64> denseShape, Operand<T> sizeOutput, Operand<U> weights, SparseBincount.Options... options) Counts the number of occurrences of each value in an integer array.<U extends TNumber>
SparseCountSparseOutput<U> SparseOps.sparseCountSparseOutput(Operand<TInt64> indices, Operand<? extends TNumber> values, Operand<TInt64> denseShape, Operand<U> weights, Boolean binaryOutput, SparseCountSparseOutput.Options... options) Performs sparse-output bin counting for a sparse tensor input.SparseOps.sparseCross(Iterable<Operand<TInt64>> indices, Iterable<Operand<?>> values, Iterable<Operand<TInt64>> shapes, Iterable<Operand<?>> denseInputs, Operand<TString> sep) Generates sparse cross from a list of sparse and dense tensors.SparseOps.sparseCrossHashed(Iterable<Operand<TInt64>> indices, Iterable<Operand<?>> values, Iterable<Operand<TInt64>> shapes, Iterable<Operand<?>> denseInputs, Operand<TInt64> numBuckets, Operand<TBool> strongHash, Operand<TInt64> salt) Generates sparse cross from a list of sparse and dense tensors.<T extends TType>
SparseDenseCwiseAdd<T> SparseOps.sparseDenseCwiseAdd(Operand<TInt64> spIndices, Operand<T> spValues, Operand<TInt64> spShape, Operand<T> dense) Adds up a SparseTensor and a dense Tensor, using these special rules: (1) Broadcasts the dense side to have the same shape as the sparse side, if eligible; (2) Then, only the dense values pointed to by the indices of the SparseTensor participate in the cwise addition.<T extends TType>
SparseDenseCwiseDiv<T> SparseOps.sparseDenseCwiseDiv(Operand<TInt64> spIndices, Operand<T> spValues, Operand<TInt64> spShape, Operand<T> dense) Component-wise divides a SparseTensor by a dense Tensor.<T extends TType>
SparseDenseCwiseMul<T> SparseOps.sparseDenseCwiseMul(Operand<TInt64> spIndices, Operand<T> spValues, Operand<TInt64> spShape, Operand<T> dense) Component-wise multiplies a SparseTensor by a dense Tensor.<T extends TType>
SparseFillEmptyRows<T> SparseOps.sparseFillEmptyRows(Operand<TInt64> indices, Operand<T> values, Operand<TInt64> denseShape, Operand<T> defaultValue) Fills empty rows in the input 2-DSparseTensorwith a default value.<T extends TType>
SparseFillEmptyRowsGrad<T> SparseOps.sparseFillEmptyRowsGrad(Operand<TInt64> reverseIndexMap, Operand<T> gradValues) The gradient of SparseFillEmptyRows.SparseOps.sparseMatMul(Operand<? extends TNumber> a, Operand<? extends TNumber> b, SparseMatMul.Options... options) Multiply matrix "a" by matrix "b".<T extends TType>
SparseMatrixAddLinalgSparseOps.sparseMatrixAdd(Operand<? extends TType> a, Operand<? extends TType> b, Operand<T> alpha, Operand<T> beta) Sparse addition of two CSR matrices, C = alpha * A + beta * B.<T extends TType>
SparseMatrixMatMul<T> LinalgSparseOps.sparseMatrixMatMul(Operand<? extends TType> a, Operand<T> b, SparseMatrixMatMul.Options... options) Matrix-multiplies a sparse matrix with a dense matrix.LinalgSparseOps.sparseMatrixMul(Operand<? extends TType> a, Operand<? extends TType> b) Element-wise multiplication of a sparse matrix with a dense tensor.LinalgSparseOps.sparseMatrixNNZ(Operand<? extends TType> sparseMatrix) Returns the number of nonzeroes ofsparse_matrix.LinalgSparseOps.sparseMatrixOrderingAMD(Operand<? extends TType> input) Computes the Approximate Minimum Degree (AMD) ordering ofinput.<T extends TNumber>
SparseMatrixSoftmaxLinalgSparseOps.sparseMatrixSoftmax(Operand<? extends TType> logits, Class<T> type) Calculates the softmax of a CSRSparseMatrix.<T extends TNumber>
SparseMatrixSoftmaxGradLinalgSparseOps.sparseMatrixSoftmaxGrad(Operand<? extends TType> softmax, Operand<? extends TType> gradSoftmax, Class<T> type) Calculates the gradient of the SparseMatrixSoftmax op.<T extends TType>
SparseMatrixSparseCholeskyLinalgSparseOps.sparseMatrixSparseCholesky(Operand<? extends TType> input, Operand<TInt32> permutation, Class<T> type) Computes the sparse Cholesky decomposition ofinput.<T extends TType>
SparseMatrixSparseMatMulLinalgSparseOps.sparseMatrixSparseMatMul(Operand<? extends TType> a, Operand<? extends TType> b, Class<T> type, SparseMatrixSparseMatMul.Options... options) Sparse-matrix-multiplies two CSR matricesaandb.<T extends TType>
SparseMatrixTransposeLinalgSparseOps.sparseMatrixTranspose(Operand<? extends TType> input, Class<T> type, SparseMatrixTranspose.Options... options) Transposes the inner (matrix) dimensions of a CSRSparseMatrix.<T extends TType>
SparseMatrixZerosLinalgSparseOps.sparseMatrixZeros(Operand<TInt64> denseShape, Class<T> type) Creates an all-zeros CSRSparseMatrix with shapedense_shape.<T extends TNumber>
SparseReduceMax<T> SparseOps.sparseReduceMax(Operand<TInt64> inputIndices, Operand<T> inputValues, Operand<TInt64> inputShape, Operand<TInt32> reductionAxes, SparseReduceMax.Options... options) Computes the max of elements across dimensions of a SparseTensor.<T extends TNumber>
SparseReduceMaxSparse<T> SparseOps.sparseReduceMaxSparse(Operand<TInt64> inputIndices, Operand<T> inputValues, Operand<TInt64> inputShape, Operand<TInt32> reductionAxes, SparseReduceMaxSparse.Options... options) Computes the max of elements across dimensions of a SparseTensor.<T extends TType>
SparseReduceSum<T> SparseOps.sparseReduceSum(Operand<TInt64> inputIndices, Operand<T> inputValues, Operand<TInt64> inputShape, Operand<TInt32> reductionAxes, SparseReduceSum.Options... options) Computes the sum of elements across dimensions of a SparseTensor.<T extends TType>
SparseReduceSumSparse<T> SparseOps.sparseReduceSumSparse(Operand<TInt64> inputIndices, Operand<T> inputValues, Operand<TInt64> inputShape, Operand<TInt32> reductionAxes, SparseReduceSumSparse.Options... options) Computes the sum of elements across dimensions of a SparseTensor.<T extends TType>
SparseReorder<T> SparseOps.sparseReorder(Operand<TInt64> inputIndices, Operand<T> inputValues, Operand<TInt64> inputShape) Reorders a SparseTensor into the canonical, row-major ordering.SparseOps.sparseReshape(Operand<TInt64> inputIndices, Operand<TInt64> inputShape, Operand<TInt64> newShape) Reshapes a SparseTensor to represent values in a new dense shape.<T extends TNumber>
SparseSegmentMean<T> SparseOps.sparseSegmentMean(Operand<T> data, Operand<? extends TNumber> indices, Operand<? extends TNumber> segmentIds, SparseSegmentMean.Options... options) Computes the mean along sparse segments of a tensor.<T extends TNumber, U extends TNumber>
SparseSegmentMeanGrad<T, U> SparseOps.sparseSegmentMeanGrad(Operand<T> grad, Operand<U> indices, Operand<? extends TNumber> segmentIds, Operand<TInt32> denseOutputDim0) Computes gradients for SparseSegmentMean.<T extends TNumber>
SparseSegmentMeanWithNumSegments<T> SparseOps.sparseSegmentMeanWithNumSegments(Operand<T> data, Operand<? extends TNumber> indices, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments, SparseSegmentMeanWithNumSegments.Options... options) Computes the mean along sparse segments of a tensor.<T extends TNumber>
SparseSegmentSqrtN<T> SparseOps.sparseSegmentSqrtN(Operand<T> data, Operand<? extends TNumber> indices, Operand<? extends TNumber> segmentIds, SparseSegmentSqrtN.Options... options) Computes the sum along sparse segments of a tensor divided by the sqrt of N.<T extends TNumber, U extends TNumber>
SparseSegmentSqrtNGrad<T, U> SparseOps.sparseSegmentSqrtNGrad(Operand<T> grad, Operand<U> indices, Operand<? extends TNumber> segmentIds, Operand<TInt32> denseOutputDim0) Computes gradients for SparseSegmentSqrtN.<T extends TNumber>
SparseSegmentSqrtNWithNumSegments<T> SparseOps.sparseSegmentSqrtNWithNumSegments(Operand<T> data, Operand<? extends TNumber> indices, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments, SparseSegmentSqrtNWithNumSegments.Options... options) Computes the sum along sparse segments of a tensor divided by the sqrt of N.<T extends TNumber>
SparseSegmentSum<T> SparseOps.sparseSegmentSum(Operand<T> data, Operand<? extends TNumber> indices, Operand<? extends TNumber> segmentIds, SparseSegmentSum.Options... options) Computes the sum along sparse segments of a tensor.<T extends TNumber, U extends TNumber>
SparseSegmentSumGrad<T, U> SparseOps.sparseSegmentSumGrad(Operand<T> grad, Operand<U> indices, Operand<? extends TNumber> segmentIds, Operand<TInt32> denseOutputDim0) Computes gradients for SparseSegmentSum.<T extends TNumber>
SparseSegmentSumWithNumSegments<T> SparseOps.sparseSegmentSumWithNumSegments(Operand<T> data, Operand<? extends TNumber> indices, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments, SparseSegmentSumWithNumSegments.Options... options) Computes the sum along sparse segments of a tensor.<T extends TType>
SparseSlice<T> SparseOps.sparseSlice(Operand<TInt64> indices, Operand<T> values, Operand<TInt64> shape, Operand<TInt64> start, Operand<TInt64> sizeOutput) Slice aSparseTensorbased on thestartandsize.<T extends TType>
SparseSliceGrad<T> SparseOps.sparseSliceGrad(Operand<T> backpropValGrad, Operand<TInt64> inputIndices, Operand<TInt64> inputStart, Operand<TInt64> outputIndices) The gradient operator for the SparseSlice op.<T extends TNumber>
SparseSoftmax<T> Applies softmax to a batched N-DSparseTensor.<T extends TNumber>
SparseSoftmaxCrossEntropyWithLogits<T> NnOps.sparseSoftmaxCrossEntropyWithLogits(Operand<T> features, Operand<? extends TNumber> labels) Computes softmax cross entropy cost and gradients to backpropagate.<T extends TNumber>
SparseSparseMaximum<T> SparseOps.sparseSparseMaximum(Operand<TInt64> aIndices, Operand<T> aValues, Operand<TInt64> aShape, Operand<TInt64> bIndices, Operand<T> bValues, Operand<TInt64> bShape) Returns the element-wise max of two SparseTensors.<T extends TType>
SparseSparseMinimum<T> SparseOps.sparseSparseMinimum(Operand<TInt64> aIndices, Operand<T> aValues, Operand<TInt64> aShape, Operand<TInt64> bIndices, Operand<T> bValues, Operand<TInt64> bShape) Returns the element-wise min of two SparseTensors.<T extends TType>
SparseSplit<T> SparseOps.sparseSplit(Operand<TInt64> splitDim, Operand<TInt64> indices, Operand<T> values, Operand<TInt64> shape, Long numSplit) Split aSparseTensorintonum_splittensors along one dimension.<U extends TType, T extends TNumber>
SparseTensorDenseAdd<U> SparseOps.sparseTensorDenseAdd(Operand<T> aIndices, Operand<U> aValues, Operand<T> aShape, Operand<U> b) Adds up aSparseTensorand a denseTensor, producing a denseTensor.<U extends TType>
SparseTensorDenseMatMul<U> SparseOps.sparseTensorDenseMatMul(Operand<? extends TNumber> aIndices, Operand<U> aValues, Operand<TInt64> aShape, Operand<U> b, SparseTensorDenseMatMul.Options... options) Multiply SparseTensor (of rank 2) "A" by dense matrix "B".DataOps.sparseTensorSliceDataset(Operand<TInt64> indices, Operand<? extends TType> values, Operand<TInt64> denseShape) Creates a dataset that splits a SparseTensor into elements row-wise.LinalgSparseOps.sparseTensorToCSRSparseMatrix(Operand<TInt64> indices, Operand<? extends TType> values, Operand<TInt64> denseShape) Converts a SparseTensor to a (possibly batched) CSRSparseMatrix.<U extends TType, T extends TNumber>
SparseToDense<U> SparseOps.sparseToDense(Operand<T> sparseIndices, Operand<T> outputShape, Operand<U> sparseValues, Operand<U> defaultValue, SparseToDense.Options... options) Converts a sparse representation into a dense tensor.<T extends TType>
SparseToSparseSetOperation<T> SparseOps.sparseToSparseSetOperation(Operand<TInt64> set1Indices, Operand<T> set1Values, Operand<TInt64> set1Shape, Operand<TInt64> set2Indices, Operand<T> set2Values, Operand<TInt64> set2Shape, String setOperation, SparseToSparseSetOperation.Options... options) Applies set operation along last dimension of 2SparseTensorinputs.The Spence operationSplits a tensor intonum_splittensors along one dimension.<T extends TNumber, U extends TNumber>
SplitDedupData<T, U> TpuOps.splitDedupData(Operand<? extends TType> input, Class<T> integerType, Class<U> floatType, String tupleMask, SplitDedupData.Options... options) An op splits input deduplication data XLA tuple into integer and floating point tensors.Splits input tensor across all dimensions.Ops.splitV(Operand<T> value, Operand<? extends TNumber> sizeSplits, Operand<TInt32> axis, Long numSplit) Splits a tensor intonum_splittensors along one dimension.DataExperimentalOps.sqlDataset(Operand<TString> driverName, Operand<TString> dataSourceName, Operand<TString> query, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that executes a SQL query and emits rows of the result set.DataOps.sqlDataset(Operand<TString> driverName, Operand<TString> dataSourceName, Operand<TString> query, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that executes a SQL query and emits rows of the result set.Computes square root of x element-wise.Computes the gradient for the sqrt ofxwrt its input.Computes the matrix square root of one or more square matrices: matmul(sqrtm(A), sqrtm(A)) = AComputes square of x element-wise.<T extends TType>
SquaredDifference<T> MathOps.squaredDifference(Operand<T> x, Operand<T> y) Returns conj(x - y)(x - y) element-wise.Ops.squeeze(Operand<T> input, Squeeze.Options... options) Removes dimensions of size 1 from the shape of a tensor.Ops.stackClose(Operand<? extends TType> handle) Delete the stack from its resource container.<T extends TType>
StackCreateOps.stackCreate(Operand<TInt32> maxSize, Class<T> elemType, StackCreate.Options... options) A stack that produces elements in first-in last-out order.Pop the element at the top of the stack.Ops.stackPush(Operand<? extends TType> handle, Operand<T> elem, StackPush.Options... options) Push an element onto the stack.Ops.stagePeek(Operand<TInt32> index, List<Class<? extends TType>> dtypes, StagePeek.Options... options) Op peeks at the values at the specified index.Ops.statefulCase(Operand<TInt32> branchIndex, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, List<ConcreteFunction> branches, Case.Options... options) An n-way switch statement which calls a single branch function.Ops.statefulIf(Operand<? extends TType> cond, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, ConcreteFunction thenBranch, ConcreteFunction elseBranch, If.Options... options) output = cond ?<U extends TNumber>
StatefulRandomBinomial<TInt64> RandomOps.statefulRandomBinomial(Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TNumber> shape, Operand<U> counts, Operand<U> probs) The StatefulRandomBinomial operation<V extends TNumber, U extends TNumber>
StatefulRandomBinomial<V> RandomOps.statefulRandomBinomial(Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TNumber> shape, Operand<U> counts, Operand<U> probs, Class<V> dtype) The StatefulRandomBinomial operationRandomOps.statefulStandardNormal(Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TType> shape) Outputs random values from a normal distribution.<U extends TType>
StatefulStandardNormal<U> RandomOps.statefulStandardNormal(Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TType> shape, Class<U> dtype) Outputs random values from a normal distribution.RandomOps.statefulTruncatedNormal(Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TType> shape) Outputs random values from a truncated normal distribution.<U extends TType>
StatefulTruncatedNormal<U> RandomOps.statefulTruncatedNormal(Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TType> shape, Class<U> dtype) Outputs random values from a truncated normal distribution.RandomOps.statefulUniform(Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TType> shape) Outputs random values from a uniform distribution.<U extends TType>
StatefulUniform<U> RandomOps.statefulUniform(Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TType> shape, Class<U> dtype) Outputs random values from a uniform distribution.<U extends TType>
StatefulUniformFullInt<U> RandomOps.statefulUniformFullInt(Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TType> shape, Class<U> dtype) Outputs random integers from a uniform distribution.<U extends TType>
StatefulUniformInt<U> RandomOps.statefulUniformInt(Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TType> shape, Operand<U> minval, Operand<U> maxval) Outputs random integers from a uniform distribution.Ops.statelessCase(Operand<TInt32> branchIndex, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, List<ConcreteFunction> branches, Case.Options... options) An n-way switch statement which calls a single branch function.Ops.statelessIf(Operand<? extends TType> cond, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, ConcreteFunction thenBranch, ConcreteFunction elseBranch, If.Options... options) output = cond ?RandomOps.statelessMultinomial(Operand<? extends TNumber> logits, Operand<TInt32> numSamples, Operand<? extends TNumber> seed) Draws samples from a multinomial distribution.<V extends TNumber>
StatelessMultinomial<V> RandomOps.statelessMultinomial(Operand<? extends TNumber> logits, Operand<TInt32> numSamples, Operand<? extends TNumber> seed, Class<V> outputDtype) Draws samples from a multinomial distribution.<V extends TNumber>
StatelessParameterizedTruncatedNormal<V> RandomOps.statelessParameterizedTruncatedNormal(Operand<? extends TNumber> shape, Operand<? extends TNumber> seed, Operand<V> means, Operand<V> stddevs, Operand<V> minvals, Operand<V> maxvals) The StatelessParameterizedTruncatedNormal operation<V extends TNumber>
StatelessRandomBinomial<TInt64> RandomOps.statelessRandomBinomial(Operand<? extends TNumber> shape, Operand<? extends TNumber> seed, Operand<V> counts, Operand<V> probs) Outputs deterministic pseudorandom random numbers from a binomial distribution.<W extends TNumber, V extends TNumber>
StatelessRandomBinomial<W> RandomOps.statelessRandomBinomial(Operand<? extends TNumber> shape, Operand<? extends TNumber> seed, Operand<V> counts, Operand<V> probs, Class<W> dtype) Outputs deterministic pseudorandom random numbers from a binomial distribution.<U extends TNumber>
StatelessRandomGamma<U> RandomOps.statelessRandomGamma(Operand<? extends TNumber> shape, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg, Operand<U> alpha) Outputs deterministic pseudorandom random numbers from a gamma distribution.RandomOps.statelessRandomGetKeyCounter(Operand<? extends TNumber> seed) Scrambles seed into key and counter, using the best algorithm based on device.RandomOps.statelessRandomGetKeyCounterAlg(Operand<? extends TNumber> seed) Picks the best algorithm based on device, and scrambles seed into key and counter.RandomOps.statelessRandomNormal(Operand<? extends TNumber> shape, Operand<? extends TNumber> seed) Outputs deterministic pseudorandom values from a normal distribution.<V extends TNumber>
StatelessRandomNormal<V> RandomOps.statelessRandomNormal(Operand<? extends TNumber> shape, Operand<? extends TNumber> seed, Class<V> dtype) Outputs deterministic pseudorandom values from a normal distribution.RandomOps.statelessRandomNormalV2(Operand<? extends TNumber> shape, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg) Outputs deterministic pseudorandom values from a normal distribution.<U extends TNumber>
StatelessRandomNormalV2<U> RandomOps.statelessRandomNormalV2(Operand<? extends TNumber> shape, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg, Class<U> dtype) Outputs deterministic pseudorandom values from a normal distribution.<W extends TNumber>
StatelessRandomPoisson<W> RandomOps.statelessRandomPoisson(Operand<? extends TNumber> shape, Operand<? extends TNumber> seed, Operand<? extends TNumber> lam, Class<W> dtype) Outputs deterministic pseudorandom random numbers from a Poisson distribution.RandomOps.statelessRandomUniform(Operand<? extends TNumber> shape, Operand<? extends TNumber> seed) Outputs deterministic pseudorandom random values from a uniform distribution.<V extends TNumber>
StatelessRandomUniform<V> RandomOps.statelessRandomUniform(Operand<? extends TNumber> shape, Operand<? extends TNumber> seed, Class<V> dtype) Outputs deterministic pseudorandom random values from a uniform distribution.<V extends TNumber>
StatelessRandomUniformFullInt<V> RandomOps.statelessRandomUniformFullInt(Operand<? extends TNumber> shape, Operand<? extends TNumber> seed, Class<V> dtype) Outputs deterministic pseudorandom random integers from a uniform distribution.<U extends TNumber>
StatelessRandomUniformFullIntV2<U> RandomOps.statelessRandomUniformFullIntV2(Operand<? extends TNumber> shape, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg, Class<U> dtype) Outputs deterministic pseudorandom random integers from a uniform distribution.<V extends TNumber>
StatelessRandomUniformInt<V> RandomOps.statelessRandomUniformInt(Operand<? extends TNumber> shape, Operand<? extends TNumber> seed, Operand<V> minval, Operand<V> maxval) Outputs deterministic pseudorandom random integers from a uniform distribution.<U extends TNumber>
StatelessRandomUniformIntV2<U> RandomOps.statelessRandomUniformIntV2(Operand<? extends TNumber> shape, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg, Operand<U> minval, Operand<U> maxval) Outputs deterministic pseudorandom random integers from a uniform distribution.RandomOps.statelessRandomUniformV2(Operand<? extends TNumber> shape, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg) Outputs deterministic pseudorandom random values from a uniform distribution.<U extends TNumber>
StatelessRandomUniformV2<U> RandomOps.statelessRandomUniformV2(Operand<? extends TNumber> shape, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg, Class<U> dtype) Outputs deterministic pseudorandom random values from a uniform distribution.<T extends TNumber>
StatelessSampleDistortedBoundingBox<T> ImageOps.statelessSampleDistortedBoundingBox(Operand<T> imageSize, Operand<TFloat32> boundingBoxes, Operand<TFloat32> minObjectCovered, Operand<? extends TNumber> seed, StatelessSampleDistortedBoundingBox.Options... options) Generate a randomly distorted bounding box for an image deterministically.<T extends TType>
StatelessShuffle<T> RandomExperimentalOps.statelessShuffle(Operand<T> value, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg) Randomly and deterministically shuffles a tensor along its first dimension.RandomOps.statelessTruncatedNormal(Operand<? extends TNumber> shape, Operand<? extends TNumber> seed) Outputs deterministic pseudorandom values from a truncated normal distribution.<V extends TNumber>
StatelessTruncatedNormal<V> RandomOps.statelessTruncatedNormal(Operand<? extends TNumber> shape, Operand<? extends TNumber> seed, Class<V> dtype) Outputs deterministic pseudorandom values from a truncated normal distribution.RandomOps.statelessTruncatedNormalV2(Operand<? extends TNumber> shape, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg) Outputs deterministic pseudorandom values from a truncated normal distribution.<U extends TNumber>
StatelessTruncatedNormalV2<U> RandomOps.statelessTruncatedNormalV2(Operand<? extends TNumber> shape, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg, Class<U> dtype) Outputs deterministic pseudorandom values from a truncated normal distribution.StringsOps.staticRegexFullMatch(Operand<TString> input, String pattern) Check if the input matches the regex pattern.StringsOps.staticRegexReplace(Operand<TString> input, String pattern, String rewrite, StaticRegexReplace.Options... options) Replaces the match of pattern in input with rewrite.DataOps.statsAggregatorSetSummaryWriter(Operand<? extends TType> statsAggregator, Operand<? extends TType> summary) Set a summary_writer_interface to record statistics using given stats_aggregator.DataExperimentalOps.statsAggregatorSummary(Operand<? extends TType> iterator) Produces a summary of any statistics recorded by the given statistics manager.SummaryOps.statsAggregatorSummary(Operand<? extends TType> iterator) Produces a summary of any statistics recorded by the given statistics manager.<U extends TNumber>
StochasticCastToInt<U> Ops.stochasticCastToInt(Operand<? extends TNumber> input, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg, Class<U> Tout) Stochastically cast a given tensor from floats to ints.<T extends TType>
StopGradient<T> Ops.stopGradient(Operand<T> input) Stops gradient computation.TpuOps.storeMinibatchStatisticsInFdo(Operand<TString> programKey, Operand<TInt32> maxIds, Operand<TInt32> maxUniques, Long sampleCount, Long numReplica, Long featureWidth, Long numScPerChip, String tableName, String miniBatchSplits) The StoreMinibatchStatisticsInFdo operation<T extends TType>
StridedSlice<T> Ops.stridedSlice(Operand<T> input, Index... indices) Return a strided slice from `input`.<T extends TType, U extends TNumber>
StridedSlice<T> Ops.stridedSlice(Operand<T> input, Operand<U> begin, Operand<U> end, Operand<U> strides, StridedSlice.Options... options) Return a strided slice frominput.<T extends TType>
StridedSliceAssign<T> Ops.stridedSliceAssign(Operand<T> ref, Operand<T> value, Index... indices) Assign `value` to the sliced l-value reference of `ref`.<T extends TType, U extends TNumber>
StridedSliceAssign<T> Ops.stridedSliceAssign(Operand<T> ref, Operand<U> begin, Operand<U> end, Operand<U> strides, Operand<T> value, StridedSliceAssign.Options... options) Assignvalueto the sliced l-value reference ofref.<U extends TType, T extends TNumber>
StridedSliceGrad<U> Ops.stridedSliceGrad(Operand<T> shape, Operand<T> begin, Operand<T> end, Operand<T> strides, Operand<U> dy, StridedSliceGrad.Options... options) Returns the gradient ofStridedSlice.StringsOps.stringLength(Operand<TString> input, StringLength.Options... options) String lengths ofinput.<T extends TNumber>
StringNGrams<T> StringsOps.stringNGrams(Operand<TString> data, Operand<T> dataSplits, String separator, List<Long> ngramWidths, String leftPad, String rightPad, Long padWidth, Boolean preserveShortSequences) Creates ngrams from ragged string data.StringsOps.stringSplit(Operand<TString> input, Operand<TString> sep, StringSplit.Options... options) Split elements ofsourcebased onsepinto aSparseTensor.Strip leading and trailing whitespaces from the Tensor.Returns x - y element-wise.StringsOps.substr(Operand<TString> input, Operand<T> pos, Operand<T> len, Substr.Options... options) Return substrings fromTensorof strings.Ops.sum(Operand<T> input, Operand<? extends TNumber> axis, Sum.Options... options) Computes the sum of elements across dimensions of a tensor.LinalgOps.svd(Operand<T> input, Svd.Options... options) Computes the singular value decompositions of one or more matrices.<T extends TType>
SwitchCond<T> Ops.switchCond(Operand<T> data, Operand<TBool> pred) Forwardsdatato the output port determined bypred.Creates a 1-dimensional operand with the dimensions matching the first n dimensions of the shape.Creates a 1-dimensional operand containin the dimensions matching the first n dimensions of the shape.DataOps.takeDataset(Operand<? extends TType> inputDataset, Operand<TInt64> count, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, TakeDataset.Options... options) Creates a dataset that containscountelements from theinput_dataset.Creates a 1-dimensional operand containing the dimensions matching the last n dimensions of the shape.Creates a 1-dimensional operand containing the dimensions matching the last n dimensions of the shape.<T extends TType>
TakeManySparseFromTensorsMap<T> SparseOps.takeManySparseFromTensorsMap(Operand<TInt64> sparseHandles, Class<T> dtype, TakeManySparseFromTensorsMap.Options... options) ReadSparseTensorsfrom aSparseTensorsMapand concatenate them.DataExperimentalOps.takeWhileDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction predicate, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that stops iteration when predicate` is false.DataOps.takeWhileDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction predicate, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, TakeWhileDataset.Options... options) Creates a dataset that stops iteration when predicate` is false.Computes tan of x element-wise.Computes hyperbolic tangent ofxelement-wise.Computes the gradient for the tanh ofxwrt its input.<T extends TType>
TensorArrayOps.tensorArray(Operand<TInt32> sizeOutput, Class<T> dtype, TensorArray.Options... options) An array of Tensors of given size.Ops.tensorArrayClose(Operand<? extends TType> handle) Delete the TensorArray from its resource container.<T extends TType>
TensorArrayConcat<T> Ops.tensorArrayConcat(Operand<? extends TType> handle, Operand<TFloat32> flowIn, Class<T> dtype, TensorArrayConcat.Options... options) Concat the elements from the TensorArray into valuevalue.<T extends TType>
TensorArrayGather<T> Ops.tensorArrayGather(Operand<? extends TType> handle, Operand<TInt32> indices, Operand<TFloat32> flowIn, Class<T> dtype, TensorArrayGather.Options... options) Gather specific elements from the TensorArray into outputvalue.Creates a TensorArray for storing the gradients of values in the given handle.Ops.tensorArrayGradWithShape(Operand<? extends TType> handle, Operand<TFloat32> flowIn, Operand<TInt32> shapeToPrepend, String source) Creates a TensorArray for storing multiple gradients of values in the given handle.<T extends TType>
TensorArrayPack<T> Ops.tensorArrayPack(Operand<TString> handle, Operand<TFloat32> flowIn, Class<T> dtype, TensorArrayPack.Options... options) The TensorArrayPack operation<T extends TType>
TensorArrayRead<T> Ops.tensorArrayRead(Operand<? extends TType> handle, Operand<TInt32> index, Operand<TFloat32> flowIn, Class<T> dtype) Read an element from the TensorArray into outputvalue.Ops.tensorArrayScatter(Operand<? extends TType> handle, Operand<TInt32> indices, Operand<? extends TType> value, Operand<TFloat32> flowIn) Scatter the data from the input value into specific TensorArray elements.Ops.tensorArraySize(Operand<? extends TType> handle, Operand<TFloat32> flowIn) Get the current size of the TensorArray.Ops.tensorArraySplit(Operand<? extends TType> handle, Operand<? extends TType> value, Operand<TInt64> lengths, Operand<TFloat32> flowIn) Split the data from the input value into TensorArray elements.Ops.tensorArrayUnpack(Operand<TString> handle, Operand<? extends TType> value, Operand<TFloat32> flowIn) The TensorArrayUnpack operationOps.tensorArrayWrite(Operand<? extends TType> handle, Operand<TInt32> index, Operand<? extends TType> value, Operand<TFloat32> flowIn) Push an element onto the tensor_array.<T extends TType>
TensorDiag<T> LinalgOps.tensorDiag(Operand<T> diagonal) Returns a diagonal tensor with a given diagonal values.<T extends TType>
TensorDiagPart<T> LinalgOps.tensorDiagPart(Operand<T> input) Returns the diagonal part of the tensor.<U extends TType>
TensorListConcat<U> Ops.tensorListConcat(Operand<? extends TType> inputHandle, Operand<? extends TNumber> elementShape, Operand<TInt64> leadingDims, Class<U> elementDtype) Concats all tensors in the list along the 0th dimension.<T extends TType>
TensorListConcatListsOps.tensorListConcatLists(Operand<? extends TType> inputA, Operand<? extends TType> inputB, Class<T> elementDtype) The TensorListConcatLists operation<T extends TNumber>
TensorListElementShape<T> Ops.tensorListElementShape(Operand<? extends TType> inputHandle, Class<T> shapeType) The shape of the elements of the given list, as a tensor.Ops.tensorListFromTensor(Operand<? extends TType> tensor, Operand<? extends TNumber> elementShape) Creates a TensorList which, when stacked, has the value oftensor.<T extends TType>
TensorListGather<T> Ops.tensorListGather(Operand<? extends TType> inputHandle, Operand<TInt32> indices, Operand<TInt32> elementShape, Class<T> elementDtype) Creates a Tensor by indexing into the TensorList.<T extends TType>
TensorListGetItem<T> Ops.tensorListGetItem(Operand<? extends TType> inputHandle, Operand<TInt32> index, Operand<TInt32> elementShape, Class<T> elementDtype) Returns the item in the list with the given index.Ops.tensorListLength(Operand<? extends TType> inputHandle) Returns the number of tensors in the input tensor list.<T extends TType>
TensorListPopBack<T> Ops.tensorListPopBack(Operand<? extends TType> inputHandle, Operand<TInt32> elementShape, Class<T> elementDtype) Returns the last element of the input list as well as a list with all but that element.Ops.tensorListPushBack(Operand<? extends TType> inputHandle, Operand<? extends TType> tensor) Returns a list which has the passed-inTensoras last element and the other elements of the given list ininput_handle.Ops.tensorListPushBackBatch(Operand<? extends TType> inputHandles, Operand<? extends TType> tensor) The TensorListPushBackBatch operation<U extends TType>
TensorListReserveOps.tensorListReserve(Operand<? extends TNumber> elementShape, Operand<TInt32> numElements, Class<U> elementDtype) List of the given size with empty elements.Ops.tensorListResize(Operand<? extends TType> inputHandle, Operand<TInt32> sizeOutput) Resizes the list.Ops.tensorListScatter(Operand<? extends TType> tensor, Operand<TInt32> indices, Operand<? extends TNumber> elementShape, Operand<TInt32> numElements) Creates a TensorList by indexing into a Tensor.Ops.tensorListScatterIntoExistingList(Operand<? extends TType> inputHandle, Operand<? extends TType> tensor, Operand<TInt32> indices) Scatters tensor at indices in an input list.Ops.tensorListSetItem(Operand<? extends TType> inputHandle, Operand<TInt32> index, Operand<? extends TType> item, TensorListSetItem.Options... options) Sets the index-th position of the list to contain the given tensor.Ops.tensorListSplit(Operand<? extends TType> tensor, Operand<? extends TNumber> elementShape, Operand<TInt64> lengths) Splits a tensor into a list.<T extends TType>
TensorListStack<T> Ops.tensorListStack(Operand<? extends TType> inputHandle, Operand<TInt32> elementShape, Class<T> elementDtype, TensorListStack.Options... options) Stacks all tensors in the list.<U extends TType>
TensorMapEraseOps.tensorMapErase(Operand<? extends TType> inputHandle, Operand<? extends TType> key, Class<U> valueDtype) Returns a tensor map with item from given key erased.Ops.tensorMapHasKey(Operand<? extends TType> inputHandle, Operand<? extends TType> key) Returns whether the given key exists in the map.Ops.tensorMapInsert(Operand<? extends TType> inputHandle, Operand<? extends TType> key, Operand<? extends TType> value) Returns a map that is the 'input_handle' with the given key-value pair inserted.<U extends TType>
TensorMapLookup<U> Ops.tensorMapLookup(Operand<? extends TType> inputHandle, Operand<? extends TType> key, Class<U> valueDtype) Returns the value from a given key in a tensor map.Ops.tensorMapSize(Operand<? extends TType> inputHandle) Returns the number of tensors in the input tensor map.<T extends TType>
TensorMapStackKeys<T> Ops.tensorMapStackKeys(Operand<? extends TType> inputHandle, Class<T> keyDtype) Returns a Tensor stack of all keys in a tensor map.<T extends TType>
TensorScatterNdAdd<T> Ops.tensorScatterNdAdd(Operand<T> tensor, Operand<? extends TNumber> indices, Operand<T> updates, TensorScatterNdAdd.Options... options) Adds sparseupdatesto an existing tensor according toindices.<T extends TType>
TensorScatterNdMax<T> Ops.tensorScatterNdMax(Operand<T> tensor, Operand<? extends TNumber> indices, Operand<T> updates, TensorScatterNdMax.Options... options) Apply a sparse update to a tensor taking the element-wise maximum.<T extends TType>
TensorScatterNdMin<T> Ops.tensorScatterNdMin(Operand<T> tensor, Operand<? extends TNumber> indices, Operand<T> updates, TensorScatterNdMin.Options... options) The TensorScatterMin operation<T extends TType>
TensorScatterNdSub<T> Ops.tensorScatterNdSub(Operand<T> tensor, Operand<? extends TNumber> indices, Operand<T> updates, TensorScatterNdSub.Options... options) Subtracts sparseupdatesfrom an existing tensor according toindices.<T extends TType>
TensorScatterNdUpdate<T> Ops.tensorScatterNdUpdate(Operand<T> tensor, Operand<? extends TNumber> indices, Operand<T> updates, TensorScatterNdUpdate.Options... options) Scatterupdatesinto an existing tensor according toindices.<T extends TType, U extends TNumber>
TensorStridedSliceUpdate<T> Ops.tensorStridedSliceUpdate(Operand<T> input, Operand<U> begin, Operand<U> end, Operand<U> strides, Operand<T> value, TensorStridedSliceUpdate.Options... options) Assignvalueto the sliced l-value reference ofinput.SummaryOps.tensorSummary(Operand<TString> tag, Operand<? extends TType> tensor, Operand<TString> serializedSummaryMetadata) Outputs aSummaryprotocol buffer with a tensor and per-plugin data.DataOps.textLineDataset(Operand<TString> filenames, Operand<TString> compressionType, Operand<TInt64> bufferSize, TextLineDataset.Options... options) Creates a dataset that emits the lines of one or more text files.DataOps.tfRecordDataset(Operand<TString> filenames, Operand<TString> compressionType, Operand<TInt64> bufferSize, Operand<TInt64> byteOffsets, TfRecordDataset.Options... options) Creates a dataset that emits the records from one or more TFRecord files.DataExperimentalOps.threadPoolDataset(Operand<? extends TType> inputDataset, Operand<? extends TType> threadPool, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that uses a custom thread pool to computeinput_dataset.DataOps.threadPoolDataset(Operand<? extends TType> inputDataset, Operand<? extends TType> threadPool, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that uses a custom thread pool to computeinput_dataset.RandomOps.threadUnsafeUnigramCandidateSampler(Operand<TInt64> trueClasses, Long numTrue, Long numSampled, Boolean unique, Long rangeMax, ThreadUnsafeUnigramCandidateSampler.Options... options) Generates labels for candidate sampling with a learned unigram distribution.Constructs a tensor by tiling a given tensor.Returns the gradient ofTile.Converts a tensor to a scalar predicate.StringsOps.toHashBucket(Operand<TString> stringTensor, Long numBuckets) Converts each string in the input Tensor to its hash mod by a number of buckets.StringsOps.toHashBucketFast(Operand<TString> input, Long numBuckets) Converts each string in the input Tensor to its hash mod by a number of buckets.Converts each string in the input Tensor to its hash mod by a number of buckets.Converts each string in the input Tensor to the specified numeric type.Converts each string in the input Tensor to the specified numeric type.NnOps.topK(Operand<T> input, Operand<? extends TNumber> k, Class<V> indexType, TopK.Options... options) Finds values and indices of theklargest elements for the last dimension.NnOps.topK(Operand<T> input, Operand<? extends TNumber> k, TopK.Options[] options) Finds values and indices of theklargest elements for the last dimension.Ops.topKUnique(Operand<TFloat32> input, Long k) Returns the TopK unique values in the array in sorted order.Ops.topKWithUnique(Operand<TFloat32> input, Long k) Returns the TopK values in the array in sorted order.TpuOps.tPUEmbeddingActivations(Operand<TFloat32> embeddingVariable, Operand<TFloat32> slicedActivations, Long tableId, Long lookupId) Deprecated.useEmbeddingActivationsinsteadTpuOps.tpuHandleToProtoKey(Operand<TInt64> uid) Converts XRT's uid handles to TensorFlow-friendly input format.<T extends TType>
TPUReplicatedOutput<T> TpuOps.tPUReplicatedOutput(Operand<T> input, Long numReplicas) Deprecated.useReplicatedOutputinsteadTpuOps.tPUReshardVariables(Iterable<Operand<? extends TType>> vars, Operand<TString> newFormatKey, Operand<? extends TType> formatStateVar) Op that reshards on-device TPU variables to specified state.Shuffle dimensions of x according to a permutation.<T extends TType>
TriangularSolve<T> LinalgOps.triangularSolve(Operand<T> matrix, Operand<T> rhs, TriangularSolve.Options... options) Solves systems of linear equations with upper or lower triangular matrices by backsubstitution.<T extends TType>
TridiagonalMatMul<T> LinalgOps.tridiagonalMatMul(Operand<T> superdiag, Operand<T> maindiag, Operand<T> subdiag, Operand<T> rhs) Calculate product with tridiagonal matrix.<T extends TType>
TridiagonalSolve<T> LinalgOps.tridiagonalSolve(Operand<T> diagonals, Operand<T> rhs, TridiagonalSolve.Options... options) Solves tridiagonal systems of equations.<T extends TType>
TruncateDiv<T> MathOps.truncateDiv(Operand<T> x, Operand<T> y) Returns x / y element-wise, rounded towards zero.<U extends TNumber>
TruncatedNormal<U> RandomOps.truncatedNormal(Operand<? extends TNumber> shape, Class<U> dtype, TruncatedNormal.Options... options) Outputs random values from a truncated normal distribution.<T extends TNumber>
TruncateMod<T> MathOps.truncateMod(Operand<T> x, Operand<T> y) Returns element-wise remainder of division.Ops.unbatch(Operand<T> batchedTensor, Operand<TInt64> batchIndex, Operand<TInt64> id, Long timeoutMicros, Unbatch.Options... options) Reverses the operation of Batch for a single output Tensor.DataExperimentalOps.unbatchDataset(Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) A dataset that splits the elements of its input into multiple elements.DataOps.unbatchDataset(Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, UnbatchDataset.Options... options) A dataset that splits the elements of its input into multiple elements.<T extends TType>
UnbatchGrad<T> Ops.unbatchGrad(Operand<T> originalInput, Operand<TInt64> batchIndex, Operand<T> grad, Operand<TInt64> id, UnbatchGrad.Options... options) Gradient of Unbatch.DataOps.uncompressElement(Operand<? extends TType> compressed, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Uncompresses a compressed dataset element.<T extends TNumber>
UnicodeDecode<T> StringsOps.unicodeDecode(Operand<TString> input, String inputEncoding, Class<T> Tsplits, UnicodeDecode.Options... options) Decodes each string ininputinto a sequence of Unicode code points.StringsOps.unicodeDecode(Operand<TString> input, String inputEncoding, UnicodeDecode.Options[] options) Decodes each string ininputinto a sequence of Unicode code points.<T extends TNumber>
UnicodeDecodeWithOffsets<T> StringsOps.unicodeDecodeWithOffsets(Operand<TString> input, String inputEncoding, Class<T> Tsplits, UnicodeDecodeWithOffsets.Options... options) Decodes each string ininputinto a sequence of Unicode code points.StringsOps.unicodeDecodeWithOffsets(Operand<TString> input, String inputEncoding, UnicodeDecodeWithOffsets.Options[] options) Decodes each string ininputinto a sequence of Unicode code points.StringsOps.unicodeEncode(Operand<TInt32> inputValues, Operand<? extends TNumber> inputSplits, String outputEncoding, UnicodeEncode.Options... options) Encode a tensor of ints into unicode strings.StringsOps.unicodeScript(Operand<TInt32> input) Determine the script codes of a given tensor of Unicode integer code points.StringsOps.unicodeTranscode(Operand<TString> input, String inputEncoding, String outputEncoding, UnicodeTranscode.Options... options) Transcode the input text from a source encoding to a destination encoding.RandomOps.uniformCandidateSampler(Operand<TInt64> trueClasses, Long numTrue, Long numSampled, Boolean unique, Long rangeMax, UniformCandidateSampler.Options... options) Generates labels for candidate sampling with a uniform distribution.<U extends TNumber>
UniformDequantize<U> QuantizationOps.uniformDequantize(Operand<? extends TNumber> input, Operand<TFloat32> scales, Operand<TInt32> zeroPoints, Class<U> Tout, Long quantizationMinVal, Long quantizationMaxVal, UniformDequantize.Options... options) Perform dequantization on the quantized Tensorinput.<U extends TNumber>
UniformQuantize<U> QuantizationOps.uniformQuantize(Operand<? extends TNumber> input, Operand<TFloat32> scales, Operand<TInt32> zeroPoints, Class<U> Tout, Long quantizationMinVal, Long quantizationMaxVal, UniformQuantize.Options... options) Perform quantization on Tensorinput.<T extends TNumber>
UniformQuantizedAdd<T> MathOps.uniformQuantizedAdd(Operand<T> lhs, Operand<T> rhs, Operand<TFloat32> lhsScales, Operand<TInt32> lhsZeroPoints, Operand<TFloat32> rhsScales, Operand<TInt32> rhsZeroPoints, Operand<TFloat32> outputScales, Operand<TInt32> outputZeroPoints, Long lhsQuantizationMinVal, Long lhsQuantizationMaxVal, Long rhsQuantizationMinVal, Long rhsQuantizationMaxVal, Long outputQuantizationMinVal, Long outputQuantizationMaxVal, UniformQuantizedAdd.Options... options) Perform quantized add of quantized Tensorlhsand quantized Tensorrhsto make quantizedoutput.<T extends TNumber>
UniformQuantizedClipByValue<T> Ops.uniformQuantizedClipByValue(Operand<T> operand, Operand<T> min, Operand<T> max, Operand<TFloat32> scales, Operand<TInt32> zeroPoints, Long quantizationMinVal, Long quantizationMaxVal, UniformQuantizedClipByValue.Options... options) Perform clip by value on the quantized Tensoroperand.<U extends TNumber, T extends TNumber>
UniformQuantizedConvolution<U> NnOps.uniformQuantizedConvolution(Operand<T> lhs, Operand<T> rhs, Operand<TFloat32> lhsScales, Operand<TInt32> lhsZeroPoints, Operand<TFloat32> rhsScales, Operand<TInt32> rhsZeroPoints, Operand<TFloat32> outputScales, Operand<TInt32> outputZeroPoints, Class<U> Tout, String padding, Long lhsQuantizationMinVal, Long lhsQuantizationMaxVal, Long rhsQuantizationMinVal, Long rhsQuantizationMaxVal, Long outputQuantizationMinVal, Long outputQuantizationMaxVal, UniformQuantizedConvolution.Options... options) Perform quantized convolution of quantized Tensorlhsand quantized Tensorrhs. to make quantizedoutput.<V extends TNumber>
UniformQuantizedConvolutionHybrid<V> NnOps.uniformQuantizedConvolutionHybrid(Operand<? extends TNumber> lhs, Operand<? extends TNumber> rhs, Operand<TFloat32> rhsScales, Operand<TInt32> rhsZeroPoints, Class<V> Tout, String padding, Long rhsQuantizationMinVal, Long rhsQuantizationMaxVal, UniformQuantizedConvolutionHybrid.Options... options) Perform hybrid quantized convolution of float Tensorlhsand quantized Tensorrhs.<U extends TNumber, T extends TNumber>
UniformQuantizedDot<U> QuantizationOps.uniformQuantizedDot(Operand<T> lhs, Operand<T> rhs, Operand<TFloat32> lhsScales, Operand<TInt32> lhsZeroPoints, Operand<TFloat32> rhsScales, Operand<TInt32> rhsZeroPoints, Operand<TFloat32> outputScales, Operand<TInt32> outputZeroPoints, Class<U> Tout, Long lhsQuantizationMinVal, Long lhsQuantizationMaxVal, Long rhsQuantizationMinVal, Long rhsQuantizationMaxVal, Long outputQuantizationMinVal, Long outputQuantizationMaxVal, UniformQuantizedDot.Options... options) Perform quantized dot of quantized Tensorlhsand quantized Tensorrhsto make quantizedoutput.<V extends TNumber>
UniformQuantizedDotHybrid<V> QuantizationOps.uniformQuantizedDotHybrid(Operand<? extends TNumber> lhs, Operand<? extends TNumber> rhs, Operand<TFloat32> rhsScales, Operand<TInt32> rhsZeroPoints, Class<V> Tout, Long rhsQuantizationMinVal, Long rhsQuantizationMaxVal, UniformQuantizedDotHybrid.Options... options) Perform hybrid quantized dot of float Tensorlhsand quantized Tensorrhs.<U extends TNumber>
UniformRequantize<U> QuantizationOps.uniformRequantize(Operand<? extends TNumber> input, Operand<TFloat32> inputScales, Operand<TInt32> inputZeroPoints, Operand<TFloat32> outputScales, Operand<TInt32> outputZeroPoints, Class<U> Tout, Long inputQuantizationMinVal, Long inputQuantizationMaxVal, Long outputQuantizationMinVal, Long outputQuantizationMaxVal, UniformRequantize.Options... options) Given quantized tensorinput, requantize it with new quantization parameters.Finds unique elements along an axis of a tensor.Finds unique elements along an axis of a tensor.DataExperimentalOps.uniqueDataset(Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that contains the unique elements ofinput_dataset.DataOps.uniqueDataset(Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, UniqueDataset.Options... options) Creates a dataset that contains the unique elements ofinput_dataset.<T extends TType>
UniqueWithCounts<T, TInt32> Ops.uniqueWithCounts(Operand<T> x, Operand<? extends TNumber> axis) Finds unique elements along an axis of a tensor.<T extends TType, V extends TNumber>
UniqueWithCounts<T, V> Ops.uniqueWithCounts(Operand<T> x, Operand<? extends TNumber> axis, Class<V> outIdx) Finds unique elements along an axis of a tensor.<T extends TNumber>
UnravelIndex<T> Ops.unravelIndex(Operand<T> indices, Operand<T> dims) Converts an array of flat indices into a tuple of coordinate arrays.StringsOps.unsortedSegmentJoin(Operand<TString> inputs, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments, UnsortedSegmentJoin.Options... options) The UnsortedSegmentJoin operation<T extends TNumber>
UnsortedSegmentMax<T> MathOps.unsortedSegmentMax(Operand<T> data, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments) Computes the maximum along segments of a tensor.<T extends TNumber>
UnsortedSegmentMin<T> MathOps.unsortedSegmentMin(Operand<T> data, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments) Computes the minimum along segments of a tensor.<T extends TType>
UnsortedSegmentProd<T> MathOps.unsortedSegmentProd(Operand<T> data, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments) Computes the product along segments of a tensor.<T extends TType>
UnsortedSegmentSum<T> MathOps.unsortedSegmentSum(Operand<T> data, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments) Computes the sum along segments of a tensor.Ops.unstack(Operand<T> value, Long num, Unstack.Options... options) Unpacks a given dimension of a rank-Rtensor intonumrank-(R-1)tensors.DataOps.unwrapDatasetVariant(Operand<? extends TType> inputHandle) The UnwrapDatasetVariant operationStringsOps.upper(Operand<TString> input, Upper.Options... options) Converts all lowercase characters into their respective uppercase replacements.<T extends TType>
UpperBound<TInt32> Ops.upperBound(Operand<T> sortedInputs, Operand<T> values) Applies upper_bound(sorted_search_values, values) along each row.<U extends TNumber, T extends TType>
UpperBound<U> Ops.upperBound(Operand<T> sortedInputs, Operand<T> values, Class<U> outType) Applies upper_bound(sorted_search_values, values) along each row.Ops.variable(Operand<T> init, Variable.Options... options) Factory method to create a new Variable with its initializer.Ops.variableShape(Operand<? extends TType> input) Returns the shape of the variable pointed to byresource.<T extends TNumber>
VariableShape<T> Ops.variableShape(Operand<? extends TType> input, Class<T> outType) Returns the shape of the variable pointed to byresource.Ops.varIsInitializedOp(Operand<? extends TType> resource) Checks whether a resource handle-based variable has been initialized.Returns locations of nonzero / true values in a tensor.DataOps.windowDataset(Operand<? extends TType> inputDataset, Operand<TInt64> sizeOutput, Operand<TInt64> shift, Operand<TInt64> stride, Operand<TBool> dropRemainder, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, WindowDataset.Options... options) Combines (nests of) input elements into a dataset of (nests of) windows.TpuOps.workerHeartbeat(Operand<TString> request) Worker heartbeat op.DataOps.wrapDatasetVariant(Operand<? extends TType> inputHandle) The WrapDatasetVariant operationSummaryOps.writeAudioSummary(Operand<? extends TType> writer, Operand<TInt64> step, Operand<TString> tag, Operand<TFloat32> tensor, Operand<TFloat32> sampleRate, WriteAudioSummary.Options... options) Writes an audio summary.Writescontentsto the file at inputfilename.SummaryOps.writeGraphSummary(Operand<? extends TType> writer, Operand<TInt64> step, Operand<TString> tensor) Writes a graph summary.SummaryOps.writeHistogramSummary(Operand<? extends TType> writer, Operand<TInt64> step, Operand<TString> tag, Operand<? extends TType> values) Writes a histogram summary.SummaryOps.writeImageSummary(Operand<? extends TType> writer, Operand<TInt64> step, Operand<TString> tag, Operand<? extends TNumber> tensor, Operand<TUint8> badColor, WriteImageSummary.Options... options) Writes an image summary.SummaryOps.writeRawProtoSummary(Operand<? extends TType> writer, Operand<TInt64> step, Operand<TString> tensor) Writes a serialized proto summary.SummaryOps.writeScalarSummary(Operand<? extends TType> writer, Operand<TInt64> step, Operand<TString> tag, Operand<? extends TNumber> value) Writes a scalar summary.SummaryOps.writeSummary(Operand<? extends TType> writer, Operand<TInt64> step, Operand<? extends TType> tensor, Operand<TString> tag, Operand<TString> summaryMetadata) Writes a tensor summary.Returns 0 if x == 0, and x / y otherwise, elementwise.XlaOps.xlaSendToHost(Operand<? extends TType> input, String key) An op to send a tensor to the host.XlaOps.xlaSparseCoreAdagrad(Operand<TInt32> indices, Operand<TFloat32> gradient, Operand<TFloat32> learningRate, Operand<TFloat32> accumulator, Operand<TFloat32> embeddingTable, Long featureWidth) The XlaSparseCoreAdagrad operationXlaOps.xlaSparseCoreAdagradMomentum(Operand<TInt32> indices, Operand<TFloat32> gradient, Operand<TFloat32> learningRate, Operand<TFloat32> beta1, Operand<TFloat32> epsilon, Operand<TFloat32> accumulator, Operand<TFloat32> momentum, Operand<TFloat32> embeddingTable, Long featureWidth, Boolean useNesterov, Float beta2, Float exponent) The XlaSparseCoreAdagradMomentum operationXlaOps.xlaSparseCoreAdam(Operand<TFloat32> embeddingTable, Operand<TInt32> indices, Operand<TFloat32> gradient, Operand<TFloat32> learningRate, Operand<TFloat32> momentum, Operand<TFloat32> velocity, Operand<TFloat32> beta1, Operand<TFloat32> beta2, Operand<TFloat32> epsilon, Long featureWidth, Boolean useSumInsideSqrt) The XlaSparseCoreAdam operationXlaOps.xlaSparseCoreFtrl(Operand<TFloat32> embeddingTable, Operand<TFloat32> accumulator, Operand<TFloat32> linear, Operand<TFloat32> learningRate, Operand<TInt32> indices, Operand<TFloat32> gradient, Operand<TFloat32> beta, Operand<TFloat32> learningRatePower, Operand<TFloat32> l2RegularizationStrength, Long featureWidth, Boolean multiplyLinearByLearningRate, Float l1RegularizationStrength) The XlaSparseCoreFtrl operationXlaOps.xlaSparseCoreSgd(Operand<TInt32> indices, Operand<TFloat32> gradient, Operand<TFloat32> learningRate, Operand<TFloat32> embeddingTable, Long featureWidth) The XlaSparseCoreSgd operationXlaOps.xlaSparseDenseMatmul(Operand<TInt32> rowIds, Operand<? extends TType> colIds, Operand<TFloat32> values, Operand<? extends TType> offsets, Operand<TFloat32> embeddingTable, Long maxIdsPerPartition, Long maxUniqueIdsPerPartition, Long inputSize) The XlaSparseDenseMatmul operationXlaOps.xlaSparseDenseMatmulGradWithAdagradAndCsrInput(Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> activationGradients, Operand<TFloat32> learningRate, Operand<TFloat32> embeddingTable, Operand<TFloat32> accumulator, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, String tableName, XlaSparseDenseMatmulGradWithAdagradAndCsrInput.Options... options) The XlaSparseDenseMatmulGradWithAdagradAndCsrInput operationXlaOps.xlaSparseDenseMatmulGradWithAdagradMomentumAndCsrInput(Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> activationGradients, Operand<TFloat32> learningRate, Operand<TFloat32> embeddingTable, Operand<TFloat32> accumulator, Operand<TFloat32> momenta, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, Boolean useNesterov, Float exponent, Float beta1, Float beta2, Float epsilon, String tableName, XlaSparseDenseMatmulGradWithAdagradMomentumAndCsrInput.Options... options) The XlaSparseDenseMatmulGradWithAdagradMomentumAndCsrInput operationXlaOps.xlaSparseDenseMatmulGradWithAdamAndCsrInput(Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> activationGradients, Operand<TFloat32> learningRate, Operand<TFloat32> embeddingTable, Operand<TFloat32> momenta, Operand<TFloat32> velocity, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, Boolean useSumInsideSqrt, Float beta1, Float beta2, Float epsilon, String tableName, XlaSparseDenseMatmulGradWithAdamAndCsrInput.Options... options) The XlaSparseDenseMatmulGradWithAdamAndCsrInput operationXlaOps.xlaSparseDenseMatmulGradWithFtrlAndCsrInput(Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> activationGradients, Operand<TFloat32> learningRate, Operand<TFloat32> embeddingTable, Operand<TFloat32> accumulator, Operand<TFloat32> linear, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, Boolean multiplyLinearByLearningRate, Float beta, Float learningRatePower, Float l1RegularizationStrength, Float l2RegularizationStrength, String tableName, XlaSparseDenseMatmulGradWithFtrlAndCsrInput.Options... options) The XlaSparseDenseMatmulGradWithFtrlAndCsrInput operationXlaOps.xlaSparseDenseMatmulGradWithSgdAndCsrInput(Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> activationGradients, Operand<TFloat32> learningRate, Operand<TFloat32> embeddingTable, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, String tableName, XlaSparseDenseMatmulGradWithSgdAndCsrInput.Options... options) The XlaSparseDenseMatmulGradWithSgdAndCsrInput operationXlaOps.xlaSparseDenseMatmulWithCsrInput(Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> embeddingTable, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, Long inputSize, Float quantizationConfigLow, Float quantizationConfigHigh, Long quantizationConfigNumBuckets, String tableName) The XlaSparseDenseMatmulWithCsrInput operationReturns 0 if x == 0, and x * log1p(y) otherwise, elementwise.Returns 0 if x == 0, and x * log(y) otherwise, elementwise.Creates a zeroed tensor given its type and shape.Returns a tensor of zeros with the same shape and type as x.Compute the Hurwitz zeta function \(\zeta(x, q)\).Method parameters in org.tensorflow.op with type arguments of type OperandModifier and TypeMethodDescription<T extends TType>
AccumulateN<T> MathOps.accumulateN(Iterable<Operand<T>> inputs, Shape shape) Returns the element-wise sum of a list of tensors.Add all input tensors element wise.static Output<?>[]Ops.assertThat(Operand<TBool> condition, Iterable<Operand<?>> data, AssertThat.Options... options) Asserts that the given condition is true.XlaOps.assignVariableConcatND(Operand<? extends TType> resource, Iterable<Operand<? extends TType>> inputs, List<Long> numConcats, AssignVariableConcatND.Options... options) Concats input tensor across all dimensions.Ops.batch(Iterable<Operand<?>> inTensors, Long numBatchThreads, Long maxBatchSize, Long batchTimeoutMicros, Long gradTimeoutMicros, Batch.Options... options) Batches all input tensors nondeterministically.Ops.batchFunction(Iterable<Operand<?>> inTensors, Iterable<Operand<?>> capturedTensors, ConcreteFunction f, Long numBatchThreads, Long maxBatchSize, Long batchTimeoutMicros, List<Class<? extends TType>> Tout, BatchFunction.Options... options) Batches all the inputs tensors to the computation done by the function.Ops.call(ConcreteFunction function, Map<String, Operand<?>> arguments) Calls the function in an execution environment, adding its graph as a function if it isn't already present.Ops.caseOp(Operand<TInt32> branchIndex, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, List<ConcreteFunction> branches, Case.Options... options) An n-way switch statement which calls a single branch function.DataOps.chooseFastestBranchDataset(Operand<? extends TType> inputDataset, Operand<TInt64> ratioNumerator, Operand<TInt64> ratioDenominator, Iterable<Operand<?>> otherArguments, Long numElementsPerBranch, List<ConcreteFunction> branches, List<Long> otherArgumentsLengths, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) The ChooseFastestBranchDataset operationDataExperimentalOps.chooseFastestDataset(Iterable<Operand<? extends TType>> inputDatasets, Long numExperiments, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) The ExperimentalChooseFastestDataset operationDataOps.chooseFastestDataset(Iterable<Operand<? extends TType>> inputDatasets, Long numExperiments, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) The ChooseFastestDataset operationTpuOps.collateTPUEmbeddingMemory(Iterable<Operand<TString>> memoryConfigs) An op that merges the string-encoded memory config protos from all hosts.<T extends TNumber>
CollectiveGather<T> CollectiveOps.collectiveGather(Operand<T> input, Operand<TInt32> groupSize, Operand<TInt32> groupKey, Operand<TInt32> instanceKey, Iterable<Operand<? extends TType>> orderingToken, CollectiveGather.Options... options) Mutually accumulates multiple tensors of identical type and shape.<T extends TNumber>
CollectiveReduceScatter<T> CollectiveOps.collectiveReduceScatter(Operand<T> input, Operand<TInt32> groupSize, Operand<TInt32> groupKey, Operand<TInt32> instanceKey, Iterable<Operand<? extends TType>> orderingToken, String mergeOp, String finalOp, CollectiveReduceScatter.Options... options) Mutually reduces multiple tensors of identical type and shape and scatters the result.TpuOps.compile(Iterable<Operand<TInt64>> dynamicShapes, Iterable<Operand<?>> guaranteedConstants, Long numComputations, ConcreteFunction function, String metadata) Compiles a computations for execution on one or more TPU devices.Ops.compositeTensorVariantFromComponents(Iterable<Operand<?>> components, String metadata) Encodes anExtensionTypevalue into avariantscalar Tensor.DataOps.compressElement(Iterable<Operand<?>> components) Compresses a dataset element.Concatenates tensors along one dimension.Concats input tensor across all dimensions.<T extends TNumber>
ConcatOffset<T> Ops.concatOffset(Operand<TInt32> concatDim, Iterable<Operand<T>> shape) Computes offsets of concat inputs within its output.TpuOps.connectTPUEmbeddingHosts(Iterable<Operand<TString>> networkConfigs) An op that sets up communication between TPUEmbedding host software instances after ConfigureTPUEmbeddingHost has been called on each host.SparseOps.convertToSparseCoreCsrWrappedCooTensor(Iterable<Operand<TInt32>> sortedRowIdsList, Iterable<Operand<TInt32>> sortedColIdsList, Iterable<Operand<TFloat32>> sortedGainsList, Iterable<Operand<TInt32>> idCountsList, Operand<TInt64> splits, Long sampleCountPerSc, Long numReplica, Long maxMinibatchesPerSc, Long maxIdsPerChipPerSample, Long tableVocabSize, Long featureWidth, String tableName, Boolean allowIdDropping) The ConvertToSparseCoreCsrWrappedCooTensor operationDataExperimentalOps.cSVDataset(Operand<TString> filenames, Operand<TString> compressionType, Operand<TInt64> bufferSize, Operand<TBool> header, Operand<TString> fieldDelim, Operand<TBool> useQuoteDelim, Operand<TString> naValue, Operand<TInt64> selectCols, Iterable<Operand<?>> recordDefaults, List<Shape> outputShapes) The ExperimentalCSVDataset operationDataOps.cSVDataset(Operand<TString> filenames, Operand<TString> compressionType, Operand<TInt64> bufferSize, Operand<TBool> header, Operand<TString> fieldDelim, Operand<TBool> useQuoteDelim, Operand<TString> naValue, Operand<TInt64> selectCols, Iterable<Operand<?>> recordDefaults, Operand<TInt64> excludeCols, List<Shape> outputShapes) The CSVDatasetV2 operation<T extends TNumber>
CudnnRNNCanonicalToParams<T> NnOps.cudnnRNNCanonicalToParams(Operand<TInt32> numLayers, Operand<TInt32> numUnits, Operand<TInt32> inputSize, Iterable<Operand<T>> weights, Iterable<Operand<T>> biases, CudnnRNNCanonicalToParams.Options... options) Converts CudnnRNN params from canonical form to usable form.IoOps.decodeCsv(Operand<TString> records, Iterable<Operand<?>> recordDefaults, DecodeCsv.Options... options) Convert CSV records to tensors.DataOps.deleteMultiDeviceIterator(Operand<? extends TType> multiDeviceIterator, Iterable<Operand<? extends TType>> iterators, Operand<? extends TType> deleter) A container for an iterator resource.DataExperimentalOps.directedInterleaveDataset(Operand<? extends TType> selectorInputDataset, Iterable<Operand<? extends TType>> dataInputDatasets, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) A substitute forInterleaveDataseton a fixed list ofNdatasets.DataOps.directedInterleaveDataset(Operand<? extends TType> selectorInputDataset, Iterable<Operand<? extends TType>> dataInputDatasets, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, DirectedInterleaveDataset.Options... options) A substitute forInterleaveDataseton a fixed list ofNdatasets.TpuOps.dynamicEnqueueTPUEmbeddingArbitraryTensorBatch(Iterable<Operand<? extends TNumber>> sampleIndicesOrRowSplits, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, Operand<TInt32> deviceOrdinal, DynamicEnqueueTPUEmbeddingArbitraryTensorBatch.Options... options) Eases the porting of code that uses tf.nn.embedding_lookup_sparse().TpuOps.dynamicEnqueueTPUEmbeddingRaggedTensorBatch(Iterable<Operand<? extends TNumber>> sampleSplits, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, Operand<TInt32> deviceOrdinal, List<Long> tableIds, DynamicEnqueueTPUEmbeddingRaggedTensorBatch.Options... options) The DynamicEnqueueTPUEmbeddingRaggedTensorBatch operation<T extends TType>
DynamicStitch<T> Interleave the values from thedatatensors into a single tensor.Tensor contraction according to Einstein summation convention.Ops.encodeProto(Operand<TInt32> sizes, Iterable<Operand<?>> values, List<String> fieldNames, String messageType, EncodeProto.Options... options) The op serializes protobuf messages provided in the input tensors.TpuOps.enqueueTPUEmbeddingArbitraryTensorBatch(Iterable<Operand<? extends TNumber>> sampleIndicesOrRowSplits, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, EnqueueTPUEmbeddingArbitraryTensorBatch.Options... options) Eases the porting of code that uses tf.nn.embedding_lookup_sparse().TpuOps.enqueueTPUEmbeddingBatch(Iterable<Operand<TString>> batch, Operand<TString> modeOverride, EnqueueTPUEmbeddingBatch.Options... options) An op that enqueues a list of input batch tensors to TPUEmbedding.TpuOps.enqueueTPUEmbeddingIntegerBatch(Iterable<Operand<TInt32>> batch, Operand<TString> modeOverride, EnqueueTPUEmbeddingIntegerBatch.Options... options) An op that enqueues a list of input batch tensors to TPUEmbedding.TpuOps.enqueueTPUEmbeddingRaggedTensorBatch(Iterable<Operand<? extends TNumber>> sampleSplits, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, List<Long> tableIds, EnqueueTPUEmbeddingRaggedTensorBatch.Options... options) Eases the porting of code that uses tf.nn.embedding_lookup().TpuOps.enqueueTPUEmbeddingSparseBatch(Iterable<Operand<? extends TNumber>> sampleIndices, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, EnqueueTPUEmbeddingSparseBatch.Options... options) An op that enqueues TPUEmbedding input indices from a SparseTensor.TpuOps.enqueueTPUEmbeddingSparseTensorBatch(Iterable<Operand<? extends TNumber>> sampleIndices, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, List<Long> tableIds, EnqueueTPUEmbeddingSparseTensorBatch.Options... options) Eases the porting of code that uses tf.nn.embedding_lookup_sparse().TpuOps.execute(Iterable<Operand<?>> args, Operand<TString> key, List<Class<? extends TType>> Tresults) Op that loads and executes a TPU program on a TPU device.TpuOps.executeAndUpdateVariables(Iterable<Operand<?>> args, Operand<TString> key, List<Class<? extends TType>> Tresults, List<Long> deviceVarReadsIndices, List<Long> deviceVarUpdatesIndices) Op that executes a program with optional in-place variable updates.DataOps.filterDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction predicate, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, FilterDataset.Options... options) Creates a dataset containing elements ofinput_datasetmatchingpredicate.DataOps.flatMapDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, FlatMapDataset.Options... options) Creates a dataset that appliesfto the outputs ofinput_dataset.Ops.forOp(Operand<TInt32> start, Operand<TInt32> limit, Operand<TInt32> delta, Iterable<Operand<?>> input, ConcreteFunction body) Applies a for loop.DataOps.generatorDataset(Iterable<Operand<?>> initFuncOtherArgs, Iterable<Operand<?>> nextFuncOtherArgs, Iterable<Operand<?>> finalizeFuncOtherArgs, ConcreteFunction initFunc, ConcreteFunction nextFunc, ConcreteFunction finalizeFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, GeneratorDataset.Options... options) Creates a dataset that invokes a function to generate elements.SparseOps.getStatsFromListOfSparseCoreCooTensors(Iterable<Operand<TInt32>> rowIdsList, Iterable<Operand<TInt32>> colIdsList, Iterable<Operand<TFloat32>> gainsList, List<Long> sampleCountList, List<Long> colOffsetList, Long numReplica, Long tableVocabSize, Long featureWidth, Long numScPerChip, String tableName) The GetStatsFromListOfSparseCoreCooTensors operationOps.gradients(Iterable<? extends Operand<?>> y, Iterable<? extends Operand<?>> x, Gradients.Options... options) Adds gradients computation ops to the graph according to scope.Ops.gradients(Operand<?> y, Iterable<? extends Operand<?>> x, Gradients.Options... options) Adds operations to compute the partial derivatives of sum ofys w.r.txs, i.e.,d(y_1 + y_2 + ...)/dx_1, d(y_1 + y_2 + ...)/dx_2...DataExperimentalOps.groupByReducerDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> keyFuncOtherArguments, Iterable<Operand<?>> initFuncOtherArguments, Iterable<Operand<?>> reduceFuncOtherArguments, Iterable<Operand<?>> finalizeFuncOtherArguments, ConcreteFunction keyFunc, ConcreteFunction initFunc, ConcreteFunction reduceFunc, ConcreteFunction finalizeFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that computes a group-by oninput_dataset.DataOps.groupByReducerDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> keyFuncOtherArguments, Iterable<Operand<?>> initFuncOtherArguments, Iterable<Operand<?>> reduceFuncOtherArguments, Iterable<Operand<?>> finalizeFuncOtherArguments, ConcreteFunction keyFunc, ConcreteFunction initFunc, ConcreteFunction reduceFunc, ConcreteFunction finalizeFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that computes a group-by oninput_dataset.DataExperimentalOps.groupByWindowDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> keyFuncOtherArguments, Iterable<Operand<?>> reduceFuncOtherArguments, Iterable<Operand<?>> windowSizeFuncOtherArguments, ConcreteFunction keyFunc, ConcreteFunction reduceFunc, ConcreteFunction windowSizeFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that computes a windowed group-by oninput_dataset. // TODO(mrry): Support non-int64 keys.DataOps.groupByWindowDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> keyFuncOtherArguments, Iterable<Operand<?>> reduceFuncOtherArguments, Iterable<Operand<?>> windowSizeFuncOtherArguments, ConcreteFunction keyFunc, ConcreteFunction reduceFunc, ConcreteFunction windowSizeFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, GroupByWindowDataset.Options... options) Creates a dataset that computes a windowed group-by oninput_dataset. // TODO(mrry): Support non-int64 keys.Returns a list of tensors with the same shapes and contents as the input tensors.Ops.ifOp(Operand<? extends TType> cond, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, ConcreteFunction thenBranch, ConcreteFunction elseBranch, If.Options... options) output = cond ?DataOps.indexFlatMapDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> mapFuncOtherArgs, Iterable<Operand<?>> indexMapFuncOtherArgs, Operand<TInt64> outputCardinality, ConcreteFunction mapFunc, ConcreteFunction indexMapFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, IndexFlatMapDataset.Options... options) The IndexFlatMapDataset operationTpuOps.infeedEnqueueTuple(Iterable<Operand<?>> inputs, List<Shape> shapes, InfeedEnqueueTuple.Options... options) Feeds multiple Tensor values into the computation as an XLA tuple.DataOps.interleaveDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> cycleLength, Operand<TInt64> blockLength, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, InterleaveDataset.Options... options) Creates a dataset that appliesfto the outputs ofinput_dataset.StringsOps.join(Iterable<Operand<TString>> inputs, Join.Options... options) Joins the strings in the given list of string tensors into one tensor; with the given separator (default is an empty separator).DataOps.legacyParallelInterleaveDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> cycleLength, Operand<TInt64> blockLength, Operand<TInt64> bufferOutputElements, Operand<TInt64> prefetchInputElements, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, LegacyParallelInterleaveDataset.Options... options) Creates a dataset that appliesfto the outputs ofinput_dataset.DataOps.listDataset(Iterable<Operand<?>> tensors, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ListDataset.Options... options) Creates a dataset that emits each oftensorsonce.TpuOps.loadAllTPUEmbeddingParameters(Iterable<Operand<TFloat32>> parameters, Iterable<Operand<TFloat32>> auxiliary1, Iterable<Operand<TFloat32>> auxiliary2, Iterable<Operand<TFloat32>> auxiliary3, Iterable<Operand<TFloat32>> auxiliary4, Iterable<Operand<TFloat32>> auxiliary5, Iterable<Operand<TFloat32>> auxiliary6, Iterable<Operand<TFloat32>> auxiliary7, String config, Long numShards, Long shardId) An op that loads optimization parameters into embedding memory.DataOps.loadDataset(Operand<TString> path, Iterable<Operand<?>> readerFuncOtherArgs, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ConcreteFunction readerFunc, LoadDataset.Options... options) The LoadDataset operationDataExperimentalOps.mapAndBatchDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> batchSize, Operand<TInt64> numParallelCalls, Operand<TBool> dropRemainder, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, MapAndBatchDataset.Options... options) Creates a dataset that fuses mapping with batching.DataOps.mapAndBatchDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> batchSize, Operand<TInt64> numParallelCalls, Operand<TBool> dropRemainder, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, MapAndBatchDataset.Options... options) Creates a dataset that fuses mapping with batching.DataExperimentalOps.mapDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, MapDataset.Options... options) Creates a dataset that appliesfto the outputs ofinput_dataset.DataOps.mapDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, MapDataset.Options... options) Creates a dataset that appliesfto the outputs ofinput_dataset.Ops.mapDefun(Iterable<Operand<?>> arguments, Iterable<Operand<?>> capturedInputs, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ConcreteFunction f, MapDefun.Options... options) Maps a function on the list of tensors unpacked from arguments on dimension 0.Ops.mapStage(Operand<TInt64> key, Operand<TInt32> indices, Iterable<Operand<?>> values, List<Class<? extends TType>> dtypes, MapStage.Options... options) Stage (key, values) in the underlying container which behaves like a hashtable.Forwards the value of an available tensor frominputstooutput.SummaryOps.mergeSummary(Iterable<Operand<TString>> inputs) Merges summaries.Ops.mlirPassthroughOp(Iterable<Operand<?>> inputs, String mlirModule, List<Class<? extends TType>> Toutputs) Wraps an arbitrary MLIR computation expressed as a module with a main() function.<T extends TNumber>
NcclReduce<T> DistributeOps.ncclReduce(Iterable<Operand<T>> input, String reduction) Reducesinputfromnum_devicesusingreductionto a single device.<T extends TNumber>
NcclReduce<T> Ops.ncclReduce(Iterable<Operand<T>> input, String reduction) Deprecated.useNcclReduceinsteadDataOps.optionalFromValue(Iterable<Operand<?>> components) Constructs an Optional variant from a tuple of tensors.Ops.orderedMapStage(Operand<TInt64> key, Operand<TInt32> indices, Iterable<Operand<?>> values, List<Class<? extends TType>> dtypes, OrderedMapStage.Options... options) Stage (key, values) in the underlying container which behaves like a ordered associative container.TpuOps.outfeedEnqueueTuple(Iterable<Operand<?>> inputs) Enqueue multiple Tensor values on the computation outfeed.DataOps.paddedBatchDataset(Operand<? extends TType> inputDataset, Operand<TInt64> batchSize, Iterable<Operand<TInt64>> paddedShapes, Iterable<Operand<?>> paddingValues, Operand<TBool> dropRemainder, List<Shape> outputShapes, PaddedBatchDataset.Options... options) Creates a dataset that batches and padsbatch_sizeelements from the input.<T extends TType>
ParallelConcat<T> Ops.parallelConcat(Iterable<Operand<T>> values, Shape shape) Concatenates a list ofNtensors along the first dimension.<T extends TType>
ParallelDynamicStitch<T> Interleave the values from thedatatensors into a single tensor.DataOps.parallelFilterDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> numParallelCalls, ConcreteFunction predicate, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ParallelFilterDataset.Options... options) Creates a dataset containing elements ofinput_datasetmatchingpredicate.DataExperimentalOps.parallelInterleaveDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> cycleLength, Operand<TInt64> blockLength, Operand<TBool> sloppy, Operand<TInt64> bufferOutputElements, Operand<TInt64> prefetchInputElements, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that appliesfto the outputs ofinput_dataset.DataOps.parallelInterleaveDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> cycleLength, Operand<TInt64> blockLength, Operand<TInt64> bufferOutputElements, Operand<TInt64> prefetchInputElements, Operand<TInt64> numParallelCalls, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ParallelInterleaveDataset.Options... options) Creates a dataset that appliesfto the outputs ofinput_dataset.DataOps.parallelMapDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> numParallelCalls, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ParallelMapDataset.Options... options) Creates a dataset that appliesfto the outputs ofinput_dataset.IoOps.parseExample(Operand<TString> serialized, Operand<TString> names, Operand<TString> sparseKeys, Operand<TString> denseKeys, Operand<TString> raggedKeys, Iterable<Operand<?>> denseDefaults, Long numSparse, List<Class<? extends TType>> sparseTypes, List<Class<? extends TType>> raggedValueTypes, List<Class<? extends TNumber>> raggedSplitTypes, List<Shape> denseShapes) Transforms a vector of tf.Example protos (as strings) into typed tensors.DataExperimentalOps.parseExampleDataset(Operand<? extends TType> inputDataset, Operand<TInt64> numParallelCalls, Iterable<Operand<?>> denseDefaults, List<String> sparseKeys, List<String> denseKeys, List<Class<? extends TType>> sparseTypes, List<Shape> denseShapes, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ParseExampleDataset.Options... options) Transformsinput_datasetcontainingExampleprotos as vectors of DT_STRING into a dataset ofTensororSparseTensorobjects representing the parsed features.DataOps.parseExampleDataset(Operand<? extends TType> inputDataset, Operand<TInt64> numParallelCalls, Iterable<Operand<?>> denseDefaults, List<String> sparseKeys, List<String> denseKeys, List<Class<? extends TType>> sparseTypes, List<Shape> denseShapes, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, List<Class<? extends TType>> raggedValueTypes, List<Class<? extends TNumber>> raggedSplitTypes, ParseExampleDataset.Options... options) Transformsinput_datasetcontainingExampleprotos as vectors of DT_STRING into a dataset ofTensororSparseTensorobjects representing the parsed features.IoOps.parseSequenceExample(Operand<TString> serialized, Operand<TString> debugName, Operand<TString> contextSparseKeys, Operand<TString> contextDenseKeys, Operand<TString> contextRaggedKeys, Operand<TString> featureListSparseKeys, Operand<TString> featureListDenseKeys, Operand<TString> featureListRaggedKeys, Operand<TBool> featureListDenseMissingAssumedEmpty, Iterable<Operand<?>> contextDenseDefaults, List<Class<? extends TType>> contextSparseTypes, List<Class<? extends TType>> contextRaggedValueTypes, List<Class<? extends TNumber>> contextRaggedSplitTypes, List<Class<? extends TType>> featureListDenseTypes, List<Class<? extends TType>> featureListSparseTypes, List<Class<? extends TType>> featureListRaggedValueTypes, List<Class<? extends TNumber>> featureListRaggedSplitTypes, ParseSequenceExample.Options... options) Transforms a vector of tf.io.SequenceExample protos (as strings) into typed tensors.IoOps.parseSingleExample(Operand<TString> serialized, Iterable<Operand<?>> denseDefaults, Long numSparse, List<String> sparseKeys, List<String> denseKeys, List<Class<? extends TType>> sparseTypes, List<Shape> denseShapes) Transforms a tf.Example proto (as a string) into typed tensors.IoOps.parseSingleSequenceExample(Operand<TString> serialized, Operand<TString> featureListDenseMissingAssumedEmpty, Iterable<Operand<TString>> contextSparseKeys, Iterable<Operand<TString>> contextDenseKeys, Iterable<Operand<TString>> featureListSparseKeys, Iterable<Operand<TString>> featureListDenseKeys, Iterable<Operand<?>> contextDenseDefaults, Operand<TString> debugName, List<Class<? extends TType>> contextSparseTypes, List<Class<? extends TType>> featureListDenseTypes, List<Class<? extends TType>> featureListSparseTypes, ParseSingleSequenceExample.Options... options) Transforms a scalar brain.SequenceExample proto (as strings) into typed tensors.Ops.partitionedCall(Iterable<Operand<?>> args, List<Class<? extends TType>> Tout, ConcreteFunction f, PartitionedCall.Options... options) returnsf(inputs), wheref's body is placed and partitioned.TpuOps.partitionedCall(Iterable<Operand<?>> args, Operand<TInt32> deviceOrdinal, List<Class<? extends TType>> Tout, ConcreteFunction f, PartitionedCall.Options... options) Calls a function placed on a specified TPU device.<T extends TType>
PartitionedInput<T> TpuOps.partitionedInput(Iterable<Operand<T>> inputs, List<Long> partitionDims, PartitionedInput.Options... options) An op that groups a list of partitioned inputs together.TpuOps.prelinearizeTuple(Iterable<Operand<?>> inputs, List<Shape> shapes, PrelinearizeTuple.Options... options) An op which linearizes multiple Tensor values to an opaque variant tensor.<T extends TType>
QuantizedConcat<T> QuantizationOps.quantizedConcat(Operand<TInt32> concatDim, Iterable<Operand<T>> values, Iterable<Operand<TFloat32>> inputMins, Iterable<Operand<TFloat32>> inputMaxes) Concatenates quantized tensors along one dimension.IoOps.queueEnqueue(Operand<? extends TType> handle, Iterable<Operand<?>> components, QueueEnqueue.Options... options) Enqueues a tuple of one or more tensors in the given queue.IoOps.queueEnqueueMany(Operand<? extends TType> handle, Iterable<Operand<?>> components, QueueEnqueueMany.Options... options) Enqueues zero or more tuples of one or more tensors in the given queue.<T extends TType, U extends TNumber>
RaggedCross<T, U> RaggedOps.raggedCross(Iterable<Operand<?>> raggedValues, Iterable<Operand<?>> raggedRowSplits, Iterable<Operand<TInt64>> sparseIndices, Iterable<Operand<?>> sparseValues, Iterable<Operand<TInt64>> sparseShape, Iterable<Operand<?>> denseInputs, String inputOrder, Boolean hashedOutput, Long numBuckets, Long hashKey, Class<T> outValuesType, Class<U> outRowSplitsType) Generates a feature cross from a list of tensors, and returns it as a RaggedTensor.<T extends TNumber, U extends TType>
RaggedGather<T, U> RaggedOps.raggedGather(Iterable<Operand<T>> paramsNestedSplits, Operand<U> paramsDenseValues, Operand<? extends TNumber> indices, Long OUTPUTRAGGEDRANK) Gather ragged slices fromparamsaxis0according toindices.<U extends TType>
RaggedTensorToSparse<U> RaggedOps.raggedTensorToSparse(Iterable<Operand<? extends TNumber>> rtNestedSplits, Operand<U> rtDenseValues) Converts aRaggedTensorinto aSparseTensorwith the same values.<U extends TType>
RaggedTensorToTensor<U> RaggedOps.raggedTensorToTensor(Operand<? extends TNumber> shape, Operand<U> values, Operand<U> defaultValue, Iterable<Operand<? extends TNumber>> rowPartitionTensors, List<String> rowPartitionTypes) Create a dense tensor from a ragged tensor, possibly altering its shape.RaggedOps.raggedTensorToVariant(Iterable<Operand<? extends TNumber>> rtNestedSplits, Operand<? extends TType> rtDenseValues, Boolean batchedInput) Encodes aRaggedTensorinto avariantTensor.DataOps.reduceDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> initialState, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ReduceDataset.Options... options) Reduces the input dataset to a singleton using a reduce function.Forwards the value of an available tensor frominputstooutput.Forwards theindexth element ofinputstooutput.Ops.remoteCall(Operand<TString> target, Iterable<Operand<?>> args, List<Class<? extends TType>> Tout, ConcreteFunction f) Runs functionfon a remote device indicated bytarget.<T extends TType>
ReplicatedInput<T> TpuOps.replicatedInput(Iterable<Operand<T>> inputs, ReplicatedInput.Options... options) Connects N inputs to an N-way replicated TPU computation.TrainOps.save(Operand<TString> prefix, Operand<TString> tensorNames, Operand<TString> shapeAndSlices, Iterable<Operand<?>> tensors) Saves tensors in V2 checkpoint format.DataOps.saveDataset(Operand<? extends TType> inputDataset, Operand<TString> path, Iterable<Operand<?>> shardFuncOtherArgs, ConcreteFunction shardFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, SaveDataset.Options... options) The SaveDatasetV2 operationTrainOps.saveSlices(Operand<TString> filename, Operand<TString> tensorNames, Operand<TString> shapesAndSlices, Iterable<Operand<?>> data) Saves input tensors slices to disk.DataExperimentalOps.scanDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> initialState, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ScanDataset.Options... options) Creates a dataset successively reducesfover the elements ofinput_dataset.DataOps.scanDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> initialState, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ScanDataset.Options... options) Creates a dataset successively reducesfover the elements ofinput_dataset.TrainOps.sdcaOptimizer(Iterable<Operand<TInt64>> sparseExampleIndices, Iterable<Operand<TInt64>> sparseFeatureIndices, Iterable<Operand<TFloat32>> sparseFeatureValues, Iterable<Operand<TFloat32>> denseFeatures, Operand<TFloat32> exampleWeights, Operand<TFloat32> exampleLabels, Iterable<Operand<TInt64>> sparseIndices, Iterable<Operand<TFloat32>> sparseWeights, Iterable<Operand<TFloat32>> denseWeights, Operand<TFloat32> exampleStateData, String lossType, Float l1, Float l2, Long numLossPartitions, Long numInnerIterations, SdcaOptimizer.Options... options) Distributed version of Stochastic Dual Coordinate Ascent (SDCA) optimizer for linear models with L1 + L2 regularization.Applies L1 regularization shrink step on the parameters.TpuOps.sendTPUEmbeddingGradients(Iterable<Operand<TFloat32>> inputs, Iterable<Operand<TFloat32>> learningRates, String config, SendTPUEmbeddingGradients.Options... options) Performs gradient updates of embedding tables.Returns shape of tensors.Returns shape of tensors.DataOps.snapshotDataset(Operand<? extends TType> inputDataset, Operand<TString> path, Iterable<Operand<?>> readerFuncOtherArgs, Iterable<Operand<?>> shardFuncOtherArgs, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ConcreteFunction readerFunc, ConcreteFunction shardFunc, SnapshotDataset.Options... options) Creates a dataset that will write to / read from a snapshot.DataOps.snapshotNestedDatasetReader(Iterable<Operand<? extends TType>> inputs, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) The SnapshotNestedDatasetReader operation<T extends TType>
SparseConcat<T> SparseOps.sparseConcat(Iterable<Operand<TInt64>> indices, Iterable<Operand<T>> values, Iterable<Operand<TInt64>> shapes, Long concatDim) Concatenates a list ofSparseTensoralong the specified dimension.SparseOps.sparseCross(Iterable<Operand<TInt64>> indices, Iterable<Operand<?>> values, Iterable<Operand<TInt64>> shapes, Iterable<Operand<?>> denseInputs, Operand<TString> sep) Generates sparse cross from a list of sparse and dense tensors.SparseOps.sparseCrossHashed(Iterable<Operand<TInt64>> indices, Iterable<Operand<?>> values, Iterable<Operand<TInt64>> shapes, Iterable<Operand<?>> denseInputs, Operand<TInt64> numBuckets, Operand<TBool> strongHash, Operand<TInt64> salt) Generates sparse cross from a list of sparse and dense tensors.Ops.stack(Iterable<Operand<T>> values, Stack.Options... options) Packs a list ofNrank-Rtensors into one rank-(R+1)tensor.Ops.stage(Iterable<Operand<?>> values, Stage.Options... options) Stage values similar to a lightweight Enqueue.Ops.statefulCase(Operand<TInt32> branchIndex, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, List<ConcreteFunction> branches, Case.Options... options) An n-way switch statement which calls a single branch function.Ops.statefulIf(Operand<? extends TType> cond, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, ConcreteFunction thenBranch, ConcreteFunction elseBranch, If.Options... options) output = cond ?Ops.statefulPartitionedCall(Iterable<Operand<?>> args, List<Class<? extends TType>> Tout, ConcreteFunction f, StatefulPartitionedCall.Options... options) returnsf(inputs), wheref's body is placed and partitioned.Ops.statefulWhile(Iterable<Operand<?>> input, ConcreteFunction cond, ConcreteFunction body, While.Options... options) output = input; While (Cond(output)) { output = Body(output) }Ops.statelessCase(Operand<TInt32> branchIndex, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, List<ConcreteFunction> branches, Case.Options... options) An n-way switch statement which calls a single branch function.Ops.statelessIf(Operand<? extends TType> cond, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, ConcreteFunction thenBranch, ConcreteFunction elseBranch, If.Options... options) output = cond ?Ops.statelessWhile(Iterable<Operand<?>> input, ConcreteFunction cond, ConcreteFunction body, While.Options... options) output = input; While (Cond(output)) { output = Body(output) }StringsOps.stringFormat(Iterable<Operand<?>> inputs, StringFormat.Options... options) Formats a string template using a list of tensors.TrainOps.symbolicGradient(Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, ConcreteFunction f) Computes the gradient function for function f via backpropagation.DataExperimentalOps.takeWhileDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction predicate, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Creates a dataset that stops iteration when predicate` is false.DataOps.takeWhileDataset(Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction predicate, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, TakeWhileDataset.Options... options) Creates a dataset that stops iteration when predicate` is false.DataOps.tensorDataset(Iterable<Operand<?>> components, List<Shape> outputShapes, TensorDataset.Options... options) Creates a dataset that emitscomponentsas a tuple of tensors once.DataOps.tensorSliceDataset(Iterable<Operand<?>> components, List<Shape> outputShapes, TensorSliceDataset.Options... options) Creates a dataset that emits each dim-0 slice ofcomponentsonce.TpuOps.tPUAnnotateTensorsWithDynamicShape(Iterable<Operand<?>> tensors) The TPUAnnotateTensorsWithDynamicShape operationTpuOps.tPUCopyWithDynamicShape(Iterable<Operand<?>> tensors, Iterable<Operand<TInt32>> unpaddedSizes) Op that copies host tensor to device with dynamic shape support.<T extends TType>
TPUReplicatedInput<T> TpuOps.tPUReplicatedInput(Iterable<Operand<T>> inputs, TPUReplicatedInput.Options... options) Deprecated.useReplicatedInputinsteadTpuOps.tPUReshardVariables(Iterable<Operand<? extends TType>> vars, Operand<TString> newFormatKey, Operand<? extends TType> formatStateVar) Op that reshards on-device TPU variables to specified state.Ops.whileOp(Iterable<Operand<?>> input, ConcreteFunction cond, ConcreteFunction body, While.Options... options) output = input; While (Cond(output)) { output = Body(output) }DataOps.windowOp(Iterable<Operand<?>> inputs, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) The WindowOp operationXlaOps.xlaHostCompute(Iterable<Operand<?>> inputs, List<Class<? extends TType>> Toutputs, List<String> ancestors, List<Shape> shapes, ConcreteFunction shapeInferenceGraph, String key, XlaHostCompute.Options... options) A pseudo-op to represent host-side computation in an XLA program.DataOps.zipDataset(Iterable<Operand<? extends TType>> inputDatasets, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ZipDataset.Options... options) Creates a dataset that zips togetherinput_datasets. -
Uses of Operand in org.tensorflow.op.audio
Classes in org.tensorflow.op.audio that implement OperandModifier and TypeClassDescriptionfinal classProduces a visualization of audio data over time.final classEncode audio data using the WAV file format.final classTransforms a spectrogram into a form that's useful for speech recognition.Fields in org.tensorflow.op.audio declared as OperandModifier and TypeFieldDescriptionEncodeWav.Inputs.audio2-D with shape[length, channels].DecodeWav.Inputs.contentsThe WAV-encoded audio, usually from a file.AudioSpectrogram.Inputs.inputFloat representation of audio data.EncodeWav.Inputs.sampleRateScalar containing the sample frequency.Mfcc.Inputs.sampleRateHow many samples per second the source audio used.Mfcc.Inputs.spectrogramTypically produced by the Spectrogram op, with magnitude_squared set to true.Methods in org.tensorflow.op.audio with parameters of type OperandModifier and TypeMethodDescriptionstatic AudioSpectrogramAudioSpectrogram.create(Scope scope, Operand<TFloat32> input, Long windowSize, Long stride, AudioSpectrogram.Options... options) Factory method to create a class wrapping a new AudioSpectrogram operation.static DecodeWavDecodeWav.create(Scope scope, Operand<TString> contents, DecodeWav.Options... options) Factory method to create a class wrapping a new DecodeWav operation.static EncodeWavFactory method to create a class wrapping a new EncodeWav operation.static MfccMfcc.create(Scope scope, Operand<TFloat32> spectrogram, Operand<TInt32> sampleRate, Mfcc.Options... options) Factory method to create a class wrapping a new Mfcc operation. -
Uses of Operand in org.tensorflow.op.bitwise
Classes in org.tensorflow.op.bitwise that implement OperandModifier and TypeClassDescriptionfinal classBitwiseAnd<T extends TNumber>Elementwise computes the bitwise AND ofxandy.final classElementwise computes the bitwise OR ofxandy.final classBitwiseXor<T extends TNumber>Elementwise computes the bitwise XOR ofxandy.final classInvert (flip) each bit of supported types; for example, typeuint8value 01010101 becomes 10101010.final classElementwise computes the bitwise left-shift ofxandy.final classRightShift<T extends TNumber>Elementwise computes the bitwise right-shift ofxandy.Fields in org.tensorflow.op.bitwise declared as OperandModifier and TypeFieldDescriptionBitwiseAnd.Inputs.xThe x inputBitwiseOr.Inputs.xThe x inputBitwiseXor.Inputs.xThe x inputInvert.Inputs.xThe x inputLeftShift.Inputs.xThe x inputRightShift.Inputs.xThe x inputBitwiseAnd.Inputs.yThe y inputBitwiseOr.Inputs.yThe y inputBitwiseXor.Inputs.yThe y inputLeftShift.Inputs.yThe y inputRightShift.Inputs.yThe y inputMethods in org.tensorflow.op.bitwise with parameters of type OperandModifier and TypeMethodDescriptionstatic <T extends TNumber>
BitwiseAnd<T> Factory method to create a class wrapping a new BitwiseAnd operation.Factory method to create a class wrapping a new BitwiseOr operation.static <T extends TNumber>
BitwiseXor<T> Factory method to create a class wrapping a new BitwiseXor operation.Factory method to create a class wrapping a new Invert operation.Factory method to create a class wrapping a new LeftShift operation.static <T extends TNumber>
RightShift<T> Factory method to create a class wrapping a new RightShift operation. -
Uses of Operand in org.tensorflow.op.cluster
Classes in org.tensorflow.op.cluster that implement OperandModifier and TypeClassDescriptionfinal classReturns the index of a data point that should be added to the seed set.final classSelects num_to_sample rows of input using the KMeans++ criterion.Fields in org.tensorflow.op.cluster declared as OperandModifier and TypeFieldDescriptionKMC2ChainInitialization.Inputs.distancesVector with squared distances to the closest previously sampled cluster center for each candidate point.KmeansPlusPlusInitialization.Inputs.numRetriesPerSampleScalar.KmeansPlusPlusInitialization.Inputs.numToSampleScalar.KmeansPlusPlusInitialization.Inputs.pointsMatrix of shape (n, d).KMC2ChainInitialization.Inputs.seedScalar.KmeansPlusPlusInitialization.Inputs.seedScalar.Methods in org.tensorflow.op.cluster with parameters of type OperandModifier and TypeMethodDescriptionstatic KMC2ChainInitializationFactory method to create a class wrapping a new KMC2ChainInitialization operation.static KmeansPlusPlusInitializationKmeansPlusPlusInitialization.create(Scope scope, Operand<TFloat32> points, Operand<TInt64> numToSample, Operand<TInt64> seed, Operand<TInt64> numRetriesPerSample) Factory method to create a class wrapping a new KmeansPlusPlusInitialization operation. -
Uses of Operand in org.tensorflow.op.collective
Classes in org.tensorflow.op.collective that implement OperandModifier and TypeClassDescriptionfinal classCollectiveAllToAll<T extends TNumber>Mutually exchanges multiple tensors of identical type and shape.final classCollectiveBcastRecv<U extends TType>Receives a tensor value broadcast from another device.final classCollectiveBcastSend<T extends TType>Broadcasts a tensor value to one or more other devices.final classCollectiveGather<T extends TNumber>Mutually accumulates multiple tensors of identical type and shape.final classInitializes a group for collective operations.final classCollectivePermute<T extends TType>An Op to permute tensors across replicated TPU instances.final classCollectiveReduce<T extends TNumber>Mutually reduces multiple tensors of identical type and shape.final classCollectiveReduceScatter<T extends TNumber>Mutually reduces multiple tensors of identical type and shape and scatters the result.Fields in org.tensorflow.op.collective declared as OperandModifier and TypeFieldDescriptionCollectiveAssignGroup.Inputs.baseKeyThe baseKey inputCollectiveAllToAll.Inputs.communicatorThe communicator inputCollectiveReduce.Inputs.communicatorThe communicator inputCollectiveAssignGroup.Inputs.deviceIndexThe deviceIndex inputCollectiveAllToAll.Inputs.groupAssignmentThe groupAssignment inputCollectiveAssignGroup.Inputs.groupAssignmentThe groupAssignment inputCollectiveReduce.Inputs.groupAssignmentThe groupAssignment inputCollectiveBcastRecv.Inputs.groupKeyThe groupKey inputCollectiveBcastSend.Inputs.groupKeyThe groupKey inputCollectiveGather.Inputs.groupKeyThe groupKey inputCollectiveInitializeCommunicator.Inputs.groupKeyThe groupKey inputCollectiveReduceScatter.Inputs.groupKeyThe groupKey inputCollectiveBcastRecv.Inputs.groupSizeThe groupSize inputCollectiveBcastSend.Inputs.groupSizeThe groupSize inputCollectiveGather.Inputs.groupSizeThe groupSize inputCollectiveInitializeCommunicator.Inputs.groupSizeThe groupSize inputCollectiveReduceScatter.Inputs.groupSizeThe groupSize inputCollectiveAllToAll.Inputs.inputThe input inputCollectiveBcastSend.Inputs.inputThe input inputCollectiveGather.Inputs.inputThe input inputCollectivePermute.Inputs.inputThe local input to be permuted.CollectiveReduce.Inputs.inputThe input inputCollectiveReduceScatter.Inputs.inputThe input inputCollectiveBcastRecv.Inputs.instanceKeyThe instanceKey inputCollectiveBcastSend.Inputs.instanceKeyThe instanceKey inputCollectiveGather.Inputs.instanceKeyThe instanceKey inputCollectiveReduceScatter.Inputs.instanceKeyThe instanceKey inputCollectiveInitializeCommunicator.Inputs.rankThe rank inputCollectiveBcastRecv.Inputs.shapeThe shape inputCollectivePermute.Inputs.sourceTargetPairsA tensor with shape [num_pairs, 2].Fields in org.tensorflow.op.collective with type parameters of type OperandModifier and TypeFieldDescriptionCollectiveGather.Inputs.orderingTokenThe orderingToken inputCollectiveReduceScatter.Inputs.orderingTokenThe orderingToken inputMethods in org.tensorflow.op.collective with parameters of type OperandModifier and TypeMethodDescriptionstatic <T extends TNumber>
CollectiveAllToAll<T> CollectiveAllToAll.create(Scope scope, Operand<T> input, Operand<? extends TType> communicator, Operand<TInt32> groupAssignment, CollectiveAllToAll.Options... options) Factory method to create a class wrapping a new CollectiveAllToAllV3 operation.static CollectiveAssignGroupCollectiveAssignGroup.create(Scope scope, Operand<TInt32> groupAssignment, Operand<TInt32> deviceIndex, Operand<TInt32> baseKey) Factory method to create a class wrapping a new CollectiveAssignGroupV2 operation.static <U extends TType>
CollectiveBcastRecv<U> CollectiveBcastRecv.create(Scope scope, Operand<TInt32> groupSize, Operand<TInt32> groupKey, Operand<TInt32> instanceKey, Operand<? extends TNumber> shape, Class<U> T, CollectiveBcastRecv.Options... options) Factory method to create a class wrapping a new CollectiveBcastRecvV2 operation.static <T extends TType>
CollectiveBcastSend<T> CollectiveBcastSend.create(Scope scope, Operand<T> input, Operand<TInt32> groupSize, Operand<TInt32> groupKey, Operand<TInt32> instanceKey, CollectiveBcastSend.Options... options) Factory method to create a class wrapping a new CollectiveBcastSendV2 operation.static <T extends TNumber>
CollectiveGather<T> CollectiveGather.create(Scope scope, Operand<T> input, Operand<TInt32> groupSize, Operand<TInt32> groupKey, Operand<TInt32> instanceKey, Iterable<Operand<? extends TType>> orderingToken, CollectiveGather.Options... options) Factory method to create a class wrapping a new CollectiveGatherV2 operation.CollectiveInitializeCommunicator.create(Scope scope, Operand<TInt32> groupKey, Operand<TInt32> rank, Operand<TInt32> groupSize, CollectiveInitializeCommunicator.Options... options) Factory method to create a class wrapping a new CollectiveInitializeCommunicator operation.static <T extends TType>
CollectivePermute<T> Factory method to create a class wrapping a new CollectivePermute operation.static <T extends TNumber>
CollectiveReduce<T> CollectiveReduce.create(Scope scope, Operand<T> input, Operand<? extends TType> communicator, Operand<TInt32> groupAssignment, String reduction, CollectiveReduce.Options... options) Factory method to create a class wrapping a new CollectiveReduceV3 operation.static <T extends TNumber>
CollectiveReduceScatter<T> CollectiveReduceScatter.create(Scope scope, Operand<T> input, Operand<TInt32> groupSize, Operand<TInt32> groupKey, Operand<TInt32> instanceKey, Iterable<Operand<? extends TType>> orderingToken, String mergeOp, String finalOp, CollectiveReduceScatter.Options... options) Factory method to create a class wrapping a new CollectiveReduceScatterV2 operation.Method parameters in org.tensorflow.op.collective with type arguments of type OperandModifier and TypeMethodDescriptionstatic <T extends TNumber>
CollectiveGather<T> CollectiveGather.create(Scope scope, Operand<T> input, Operand<TInt32> groupSize, Operand<TInt32> groupKey, Operand<TInt32> instanceKey, Iterable<Operand<? extends TType>> orderingToken, CollectiveGather.Options... options) Factory method to create a class wrapping a new CollectiveGatherV2 operation.static <T extends TNumber>
CollectiveReduceScatter<T> CollectiveReduceScatter.create(Scope scope, Operand<T> input, Operand<TInt32> groupSize, Operand<TInt32> groupKey, Operand<TInt32> instanceKey, Iterable<Operand<? extends TType>> orderingToken, String mergeOp, String finalOp, CollectiveReduceScatter.Options... options) Factory method to create a class wrapping a new CollectiveReduceScatterV2 operation. -
Uses of Operand in org.tensorflow.op.core
Classes in org.tensorflow.op.core that implement OperandModifier and TypeClassDescriptionfinal classComputes the "logical and" of elements across dimensions of a tensor.final classCreates a uninitialized anonymous hash table.final classCreates an empty anonymous mutable hash table that uses tensors as the backing store.final classCreates an empty anonymous mutable hash table.final classCreates an empty anonymous mutable hash table of vector values.final classComputes the "logical or" of elements across dimensions of a tensor.final classUpdate 'ref' by assigning 'value' to it.final classUpdate 'ref' by adding 'value' to it.final classUpdate 'ref' by subtracting 'value' from it.final classDefines a barrier that persists across different graph executions.final classComputes the number of incomplete elements in the given barrier.final classComputes the number of complete elements in the given barrier.final classBatchToSpace<T extends TType>BatchToSpace for 4-D tensors of type T.final classBatchToSpaceNd<T extends TType>BatchToSpace for N-D tensors of type T.final classBitcasts a tensor from one type to another without copying data.final classBroadcastDynamicShape<T extends TNumber>Return the shape of s0 op s1 with broadcast.final classBroadcastTo<T extends TType>Broadcast an array for a compatible shape.final classBucketizes 'input' based on 'boundaries'.final classCheckPinned<T extends TType>Checks whether a tensor is located in host memory pinned for GPU.final classClipByValue<T extends TType>Clips tensor values to a specified min and max.final classEncodes anExtensionTypevalue into avariantscalar Tensor.final classConcatenates tensors along one dimension.final classAn operator producing a constant value.final classCopy a tensor from CPU-to-CPU or GPU-to-GPU.final classCopy a tensor to host.final classCopyToMesh<T extends TType>The CopyToMesh operationfinal classCopyToMeshGrad<T extends TType>The CopyToMeshGrad operationfinal classIncrements 'ref' until it reaches 'limit'.final classMakes a copy ofx.final classDestroyTemporaryVariable<T extends TType>Destroys the temporary variable and returns its final value.final classReturn the index of device the op runs.final classThe DummyMemoryCache operationfinal classDynamicStitch<T extends TType>Interleave the values from thedatatensors into a single tensor.final classComputes the (possibly normalized) Levenshtein Edit Distance.final classCreates a tensor with the given shape.final classCreates and returns an empty tensor list.final classCreates and returns an empty tensor map. handle: an empty tensor mapfinal classThe op serializes protobuf messages provided in the input tensors.final classEnsureShape<T extends TType>Ensures that the tensor's shape matches the expected shape.final classCreates or finds a child frame, and makesdataavailable to the child frame.final classExits the current frame to its parent frame.final classExpandDims<T extends TType>Inserts a dimension of 1 into a tensor's shape.final classExtractVolumePatches<T extends TNumber>Extractpatchesfrominputand put them in the"depth"output dimension. 3D extension ofextract_image_patches.final classThis op is used as a placeholder in If branch functions.final classCreates a tensor filled with a scalar value.final classGenerates fingerprint values.final classGather slices fromparamsaxisaxisaccording toindices.final classGather slices fromparamsinto a Tensor with shape specified byindices.final classReturns thetf.data.Optionsattached toinput_dataset.final classStore the input tensor in the state of the current session.final classGetSessionTensor<T extends TType>Get the value of the tensor specified by its handle.final classGuaranteeConst<T extends TType>Gives a guarantee to the TF runtime that the input tensor is a constant.final classCreates a non-initialized hash table.final classHistogramFixedWidth<U extends TNumber>Return histogram of values.final classReturns a constant tensor on the host.final classReturn a tensor with the same shape and contents as the input tensor or value.final classImmutableConst<T extends TType>Returns immutable tensor from memory region.final classInplaceAdd<T extends TType>Adds v into specified rows of x.final classInplaceSub<T extends TType>Subtracts `v` into specified rows of `x`.final classInplaceUpdate<T extends TType>Updates specified rows 'i' with values 'v'.final classChecks whether a tensor has been initialized.final classComputes the Kth order statistic of a data set.final classGenerates values in an interval.final classLookupTableFind<U extends TType>Looks up keys in a table, outputs the corresponding values.final classComputes the number of elements in the given table.final classForwards the input to the output.final classLowerBound<U extends TNumber>Applies lower_bound(sorted_search_values, values) along each row.final classMake all elements in the non-Batch dimension unique, but "close" to their initial value.final classOp returns the number of incomplete elements in the underlying container.final classOp returns the number of elements in the underlying container.final classComputes the maximum of elements across dimensions of a tensor.final classComputes the minimum of elements across dimensions of a tensor.final classPads a tensor with mirrored values.final classMirrorPadGrad<T extends TType>Gradient op forMirrorPadop.final classCreates an empty hash table that uses tensors as the backing store.final classCreates an empty hash table.final classCreates an empty hash table.final classCreates a Mutex resource that can be locked byMutexLock.final classLocks a mutex resource.final classNcclAllReduce<T extends TNumber>Deprecated.useNcclAllReduceinsteadfinal classNcclBroadcast<T extends TNumber>Deprecated.useNcclBroadcastinsteadfinal classNcclReduce<T extends TNumber>Deprecated.useNcclReduceinsteadfinal classNextIteration<T extends TType>Makes its input available to the next iteration.final classReturns a one-hot tensor.final classAn operator creating a constant initialized with ones of the shape given by `dims`.final classReturns a tensor of ones with the same shape and type as x.final classOp returns the number of incomplete elements in the underlying container.final classOp returns the number of elements in the underlying container.final classPads a tensor.final classParallelConcat<T extends TType>Concatenates a list ofNtensors along the first dimension.final classParallelDynamicStitch<T extends TType>Interleave the values from thedatatensors into a single tensor.final classPlaceholder<T extends TType>A placeholder op for a value that will be fed into the computation.final classPlaceholderWithDefault<T extends TType>A placeholder op that passes throughinputwhen its output is not fed.final classComputes the product of elements across dimensions of a tensor.final classRandomIndexShuffle<T extends TNumber>Outputs the position ofvaluein a permutation of [0, ..., max_index].final classCreates a sequence of numbers.final classReturns the rank of a tensor.final classReadVariableOp<T extends TType>Reads the value of a variable.final classReceives the named tensor from send_device on recv_device.final classComputes the "logical and" of elements across dimensions of a tensor.final classComputes the "logical or" of elements across dimensions of a tensor.final classComputes the maximum of elements across dimensions of a tensor.final classComputes the minimum of elements across dimensions of a tensor.final classReduceProd<T extends TType>Computes the product of elements across dimensions of a tensor.final classComputes the sum of elements across dimensions of a tensor.final classCreates or finds a child frame, and makesdataavailable to the child frame.final classExits the current frame to its parent frame.final classRefIdentity<T extends TType>Return the same ref tensor as the input ref tensor.final classRefNextIteration<T extends TType>Makes its input available to the next iteration.final classForwards theindexth element ofinputstooutput.final classThe Relayout operationfinal classRelayoutLike<T extends TType>The RelayoutLike operationfinal classReshapes a tensor.final classResourceCountUpTo<T extends TNumber>Increments variable pointed to by 'resource' until it reaches 'limit'.final classResourceGather<U extends TType>Gather slices from the variable pointed to byresourceaccording toindices.final classResourceGatherNd<U extends TType>The ResourceGatherNd operationfinal classReverses specific dimensions of a tensor.final classReverseSequence<T extends TType>Reverses variable length slices.final classRolls the elements of a tensor along an axis.final classScatterAdd<T extends TType>Adds sparse updates to a variable reference.final classScatterDiv<T extends TType>Divides a variable reference by sparse updates.final classScatterMax<T extends TNumber>Reduces sparse updates into a variable reference using themaxoperation.final classScatterMin<T extends TNumber>Reduces sparse updates into a variable reference using theminoperation.final classScatterMul<T extends TType>Multiplies sparse updates into a variable reference.final classScattersupdatesinto a tensor of shapeshapeaccording toindices.final classScatterNdAdd<T extends TType>Applies sparse addition to individual values or slices in a Variable.final classScatterNdMax<T extends TType>Computes element-wise maximum.final classScatterNdMin<T extends TType>Computes element-wise minimum.final classScatterNdNonAliasingAdd<T extends TType>Applies sparse addition toinputusing individual values or slices fromupdatesaccording to indicesindices.final classScatterNdSub<T extends TType>Applies sparse subtraction to individual values or slices in a Variable. within a given variable according toindices.final classScatterNdUpdate<T extends TType>Applies sparseupdatesto individual values or slices within a given variable according toindices.final classScatterSub<T extends TType>Subtracts sparse updates to a variable reference.final classScatterUpdate<T extends TType>Applies sparse updates to a variable reference.final classThe SelectV2 operationfinal classNumber of unique elements along last dimension of inputset.final classReturns the shape of a tensor.final classReturns the size of a tensor.final classReturn a slice from 'input'.final classReturns a copy of the input tensor.final classSpaceToBatchNd<T extends TType>SpaceToBatch for N-D tensors of type T.final classRemoves dimensions of size 1 from the shape of a tensor.final classPacks a list ofNrank-Rtensors into one rank-(R+1)tensor.final classA stack that produces elements in first-in last-out order.final classPop the element at the top of the stack.final classPush an element onto the stack.final classOp returns the number of elements in the underlying container.final classStochasticCastToInt<U extends TNumber>Stochastically cast a given tensor from floats to ints.final classStopGradient<T extends TType>Stops gradient computation.final classStridedSlice<T extends TType>Return a strided slice frominput.final classStridedSliceAssign<T extends TType>Assignvalueto the sliced l-value reference ofref.final classStridedSliceGrad<U extends TType>Returns the gradient ofStridedSlice.final classComputes the sum of elements across dimensions of a tensor.final classTemporaryVariable<T extends TType>Returns a tensor that may be mutated, but only persists within a single step.final classTensorArrayGather<T extends TType>Gather specific elements from the TensorArray into outputvalue.final classTensorArrayPack<T extends TType>The TensorArrayPack operationfinal classTensorArrayRead<T extends TType>Read an element from the TensorArray into outputvalue.final classScatter the data from the input value into specific TensorArray elements.final classGet the current size of the TensorArray.final classSplit the data from the input value into TensorArray elements.final classThe TensorArrayUnpack operationfinal classPush an element onto the tensor_array.final classThe TensorListConcatLists operationfinal classTensorListElementShape<T extends TNumber>The shape of the elements of the given list, as a tensor. input_handle: the list element_shape: the shape of elements of the listfinal classCreates a TensorList which, when stacked, has the value oftensor.final classTensorListGather<T extends TType>Creates a Tensor by indexing into the TensorList.final classTensorListGetItem<T extends TType>Returns the item in the list with the given index. input_handle: the list index: the position in the list from which an element will be retrieved item: the element at that positionfinal classReturns the number of tensors in the input tensor list. input_handle: the input list length: the number of tensors in the listfinal classReturns a list which has the passed-inTensoras last element and the other elements of the given list ininput_handle. tensor: The tensor to put on the list. input_handle: The old list. output_handle: A list with the elements of the old list followed by tensor. element_dtype: the type of elements in the list. element_shape: a shape compatible with that of elements in the list.final classThe TensorListPushBackBatch operationfinal classList of the given size with empty elements. element_shape: the shape of the future elements of the list num_elements: the number of elements to reserve handle: the output list element_dtype: the desired type of elements in the list.final classResizes the list. input_handle: the input list size: size of the output listfinal classCreates a TensorList by indexing into a Tensor.final classScatters tensor at indices in an input list.final classSets the index-th position of the list to contain the given tensor. input_handle: the list index: the position in the list to which the tensor will be assigned item: the element to be assigned to that position output_handle: the new list, with the element in the proper positionfinal classSplits a tensor into a list. list[i] corresponds to lengths[i] tensors from the input tensor.final classTensorListStack<T extends TType>Stacks all tensors in the list.final classReturns a tensor map with item from given key erased. input_handle: the original map output_handle: the map with value from given key removed key: the key of the value to be erasedfinal classReturns whether the given key exists in the map. input_handle: the input map key: the key to check has_key: whether the key is already in the map or notfinal classReturns a map that is the 'input_handle' with the given key-value pair inserted. input_handle: the original map output_handle: the map with key and value inserted key: the key to be inserted value: the value to be insertedfinal classTensorMapLookup<U extends TType>Returns the value from a given key in a tensor map. input_handle: the input map key: the key to be looked up value: the value found from the given keyfinal classReturns the number of tensors in the input tensor map. input_handle: the input map size: the number of tensors in the mapfinal classTensorMapStackKeys<T extends TType>Returns a Tensor stack of all keys in a tensor map. input_handle: the input map keys: the returned Tensor of all keys in the mapfinal classTensorScatterNdAdd<T extends TType>Adds sparseupdatesto an existing tensor according toindices.final classTensorScatterNdMax<T extends TType>Apply a sparse update to a tensor taking the element-wise maximum.final classTensorScatterNdMin<T extends TType>The TensorScatterMin operationfinal classTensorScatterNdSub<T extends TType>Subtracts sparseupdatesfrom an existing tensor according toindices.final classTensorScatterNdUpdate<T extends TType>Scatterupdatesinto an existing tensor according toindices.final classTensorStridedSliceUpdate<T extends TType>Assignvalueto the sliced l-value reference ofinput.final classConstructs a tensor by tiling a given tensor.final classProvides the time since epoch in seconds.final classReverses the operation of Batch for a single output Tensor.final classUnbatchGrad<T extends TType>Gradient of Unbatch.final classUniformQuantizedClipByValue<T extends TNumber>Perform clip by value on the quantized Tensoroperand.final classUnravelIndex<T extends TNumber>Converts an array of flat indices into a tuple of coordinate arrays.final classUpperBound<U extends TNumber>Applies upper_bound(sorted_search_values, values) along each row.final classCreates a handle to a Variable resource.final classHolds state in the form of a tensor that persists across steps.final classVariableShape<T extends TNumber>Returns the shape of the variable pointed to byresource.final classChecks whether a resource handle-based variable has been initialized.final classReturns locations of nonzero / true values in a tensor.final classAn operator creating a constant initialized with zeros of the shape given by `dims`.final classReturns a tensor of zeros with the same shape and type as x.Subinterfaces with type arguments of type Operand in org.tensorflow.op.coreModifier and TypeInterfaceDescriptioninterfaceAn n-way switch statement which calls a single branch function.interfaceoutput = cond ?interfaceoutput = input; While (Cond(output)) { output = Body(output) }Classes in org.tensorflow.op.core that implement interfaces with type arguments of type OperandModifier and TypeClassDescriptionfinal classBatches all the inputs tensors to the computation done by the function.final classDecodes avariantscalar Tensor into anExtensionTypevalue.final classConcatOffset<T extends TNumber>Computes offsets of concat inputs within its output.final classDynamicPartition<T extends TType>Partitionsdataintonum_partitionstensors using indices frompartitions.final classApplies a for loop.final classGets the element at the specified index in a dataset.final classAdds operations to compute the partial derivatives of sum ofys w.r.txs, i.e.,d(y_1 + y_2 + ...)/dx_1, d(y_1 + y_2 + ...)/dx_2...final classReturns a list of tensors with the same shapes and contents as the input tensors.final classMaps a function on the list of tensors unpacked from arguments on dimension 0.final classOp peeks at the values at the specified key.final classOp removes and returns the values associated with the key from the underlying container.final classWraps an arbitrary MLIR computation expressed as a module with a main() function.final classOp peeks at the values at the specified key.final classOp removes and returns the values associated with the key from the underlying container.final classreturnsf(inputs), wheref's body is placed and partitioned.final classRuns functionfon a remote device indicated bytarget.final classReturns shape of tensors.final classSplits a tensor intonum_splittensors along one dimension.final classSplits a tensor intonum_splittensors along one dimension.final classOp peeks at the values at the specified index.final classreturnsf(inputs), wheref's body is placed and partitioned.final classUnpacks a given dimension of a rank-Rtensor intonumrank-(R-1)tensors.final classOp is similar to a lightweight Dequeue.Fields in org.tensorflow.op.core declared as OperandModifier and TypeFieldDescriptionStochasticCastToInt.Inputs.algThe RNG algorithm (shape int32[]).All.Inputs.axisThe dimensions to reduce.Any.Inputs.axisThe dimensions to reduce.Concat.Inputs.axis0-D.ExpandDims.Inputs.axis0-D (scalar).Gather.Inputs.axisThe axis inparamsto gatherindicesfrom.Max.Inputs.axisThe dimensions to reduce.Min.Inputs.axisThe dimensions to reduce.Prod.Inputs.axisThe dimensions to reduce.ReduceAll.Inputs.axisThe dimensions to reduce.ReduceAny.Inputs.axisThe dimensions to reduce.ReduceMax.Inputs.axisThe dimensions to reduce.ReduceMin.Inputs.axisThe dimensions to reduce.ReduceProd.Inputs.axisThe dimensions to reduce.ReduceSum.Inputs.axisThe dimensions to reduce.Reverse.Inputs.axis1-D.Roll.Inputs.axisDimension must be 0-D or 1-D.Split.Inputs.axis0-D.SplitV.Inputs.axis0-D.Sum.Inputs.axisThe dimensions to reduce.Unique.Inputs.axisATensorof typeint32(default: None).UniqueWithCounts.Inputs.axisATensorof typeint32(default: None).Unbatch.Inputs.batchedTensorThe batchedTensor inputUnbatch.Inputs.batchIndexThe batchIndex inputUnbatchGrad.Inputs.batchIndexThe batchIndex inputResourceStridedSliceAssign.Inputs.beginThe begin inputSlice.Inputs.beginbegin[i] specifies the offset into the 'i'th dimension of 'input' to slice from.StridedSlice.Inputs.beginbegin[k]specifies the offset into thekth range specification.StridedSliceAssign.Inputs.beginThe begin inputStridedSliceGrad.Inputs.beginThe begin inputTensorStridedSliceUpdate.Inputs.beginThe begin inputBatchToSpaceNd.Inputs.blockShape1-D with shape[M], all values must be >= 1.SpaceToBatchNd.Inputs.blockShape1-D with shape[M], all values must be >= 1.StatefulCase.Inputs.branchIndexThe branch selector, an int32 Tensor.StatelessCase.Inputs.branchIndexThe branch selector, an int32 Tensor.DecodeProto.Inputs.bytesTensor of serialized protos with shapebatch_shape.ClipByValue.Inputs.clipValueMaxA 0-D (scalar)Tensor, or aTensorwith the same shape ast.ClipByValue.Inputs.clipValueMinA 0-D (scalar)Tensor, or aTensorwith the same shape ast.ConcatOffset.Inputs.concatDimThe dimension along which to concatenate.StatefulIf.Inputs.condA Tensor.StatelessIf.Inputs.condA Tensor.AssertThat.Inputs.conditionThe condition to evaluate.Select.Inputs.conditionThe condition inputWhere.Inputs.conditionThe condition inputPad.Inputs.constantValuesThe constantValues inputStochasticCastToInt.Inputs.counterInitial counter for the counter-based RNG algorithm (shape uint64[2] or uint64[1] depending on the algorithm).BatchToSpace.Inputs.crops2-D tensor of non-negative integers with shape[2, 2].BatchToSpaceNd.Inputs.crops2-D with shape[M, 2], all values must be >= 0.DynamicPartition.Inputs.dataThe data inputEnter.Inputs.dataThe tensor to be made available to the child frame.Exit.Inputs.dataThe tensor to be made available to the parent frame.Fingerprint.Inputs.dataMust have rank 1 or higher.NextIteration.Inputs.dataThe tensor to be made available to the next iteration.RefEnter.Inputs.dataThe tensor to be made available to the child frame.RefExit.Inputs.dataThe tensor to be made available to the parent frame.RefNextIteration.Inputs.dataThe tensor to be made available to the next iteration.RefSwitch.Inputs.dataThe ref tensor to be forwarded to the appropriate output.SwitchCond.Inputs.dataThe tensor to be forwarded to the appropriate output.GetElementAtIndex.Inputs.datasetThe dataset inputLookupTableFind.Inputs.defaultValueThe defaultValue inputAnonymousMutableDenseHashTable.Inputs.deletedKeyThe deletedKey inputMutableDenseHashTable.Inputs.deletedKeyThe deletedKey inputFor.Inputs.deltaThe increment.Range.Inputs.delta0-D (scalar).OneHot.Inputs.depthA scalar defining the depth of the one hot dimension.Fill.Inputs.dims1-D.UnravelIndex.Inputs.dimsAn 1-DintTensor.StridedSliceGrad.Inputs.dyThe dy inputSelect.Inputs.eThe e inputStackPush.Inputs.elemThe tensor to be pushed onto the stack.EmptyTensorList.Inputs.elementShapeThe elementShape inputTensorListConcat.Inputs.elementShapeThe elementShape inputTensorListFromTensor.Inputs.elementShapeThe elementShape inputTensorListGather.Inputs.elementShapeThe elementShape inputTensorListGetItem.Inputs.elementShapeThe elementShape inputTensorListPopBack.Inputs.elementShapeThe elementShape inputTensorListReserve.Inputs.elementShapeThe elementShape inputTensorListScatter.Inputs.elementShapeThe elementShape inputTensorListSplit.Inputs.elementShapeThe elementShape inputTensorListStack.Inputs.elementShapeThe elementShape inputAnonymousMutableDenseHashTable.Inputs.emptyKeyThe key used to represent empty key buckets internally.MutableDenseHashTable.Inputs.emptyKeyThe key used to represent empty key buckets internally.CompositeTensorVariantToComponents.Inputs.encodedA scalarvariantTensor containing an encoded ExtensionType value.ResourceStridedSliceAssign.Inputs.endThe end inputStridedSlice.Inputs.endend[i]is likebeginwith the exception thatend_maskis used to determine full ranges.StridedSliceAssign.Inputs.endThe end inputStridedSliceGrad.Inputs.endThe end inputTensorStridedSliceUpdate.Inputs.endThe end inputInitializeTableFromTextFile.Inputs.filenameFilename of a vocabulary text file.TensorArrayConcat.Inputs.flowInA float scalar that enforces proper chaining of operations.TensorArrayGather.Inputs.flowInA float scalar that enforces proper chaining of operations.TensorArrayGrad.Inputs.flowInA float scalar that enforces proper chaining of operations.TensorArrayGradWithShape.Inputs.flowInA float scalar that enforces proper chaining of operations.TensorArrayPack.Inputs.flowInThe flowIn inputTensorArrayRead.Inputs.flowInA float scalar that enforces proper chaining of operations.TensorArrayScatter.Inputs.flowInA float scalar that enforces proper chaining of operations.TensorArraySize.Inputs.flowInA float scalar that enforces proper chaining of operations.TensorArraySplit.Inputs.flowInA float scalar that enforces proper chaining of operations.TensorArrayUnpack.Inputs.flowInThe flowIn inputTensorArrayWrite.Inputs.flowInA float scalar that enforces proper chaining of operations.CopyToMeshGrad.Inputs.forwardInputThe forwardInput inputUnbatchGrad.Inputs.gradThe grad inputBarrierClose.Inputs.handleThe handle to a barrier.BarrierIncompleteSize.Inputs.handleThe handle to a barrier.BarrierInsertMany.Inputs.handleThe handle to a barrier.BarrierReadySize.Inputs.handleThe handle to a barrier.BarrierTakeMany.Inputs.handleThe handle to a barrier.DeleteSessionTensor.Inputs.handleThe handle for a tensor stored in the session state.GetSessionTensor.Inputs.handleThe handle for a tensor stored in the session state.StackClose.Inputs.handleThe handle to a stack.StackPop.Inputs.handleThe handle to a stack.StackPush.Inputs.handleThe handle to a stack.TensorArrayClose.Inputs.handleThe handle to a TensorArray (output of TensorArray or TensorArrayGrad).TensorArrayConcat.Inputs.handleThe handle to a TensorArray.TensorArrayGather.Inputs.handleThe handle to a TensorArray.TensorArrayGrad.Inputs.handleThe handle to the forward TensorArray.TensorArrayGradWithShape.Inputs.handleThe handle to the forward TensorArray.TensorArrayPack.Inputs.handleThe handle inputTensorArrayRead.Inputs.handleThe handle to a TensorArray.TensorArrayScatter.Inputs.handleThe handle to a TensorArray.TensorArraySize.Inputs.handleThe handle to a TensorArray (output of TensorArray or TensorArrayGrad).TensorArraySplit.Inputs.handleThe handle to a TensorArray.TensorArrayUnpack.Inputs.handleThe handle inputTensorArrayWrite.Inputs.handleThe handle to a TensorArray.EditDistance.Inputs.hypothesisIndicesThe indices of the hypothesis list SparseTensor.EditDistance.Inputs.hypothesisShapeThe shape of the hypothesis list SparseTensor.EditDistance.Inputs.hypothesisValuesThe values of the hypothesis list SparseTensor.InplaceAdd.Inputs.iA vector.InplaceSub.Inputs.iA vector.InplaceUpdate.Inputs.iA vector.Unbatch.Inputs.idThe id inputUnbatchGrad.Inputs.idThe id inputGetElementAtIndex.Inputs.indexThe index inputRandomIndexShuffle.Inputs.indexA scalar tensor or a vector of dtypedtype.RefSelect.Inputs.indexA scalar that determines the input that gets selected.StagePeek.Inputs.indexThe index inputTensorArrayRead.Inputs.indexThe index inputTensorArrayWrite.Inputs.indexThe position to write to inside the TensorArray.TensorListGetItem.Inputs.indexThe index inputTensorListSetItem.Inputs.indexThe index inputGather.Inputs.indicesIndex tensor.GatherNd.Inputs.indicesIndex tensor.MapPeek.Inputs.indicesThe indices inputMapStage.Inputs.indicesThe indices inputMapUnstage.Inputs.indicesThe indices inputMapUnstageNoKey.Inputs.indicesThe indices inputOneHot.Inputs.indicesA tensor of indices.OrderedMapPeek.Inputs.indicesThe indices inputOrderedMapStage.Inputs.indicesThe indices inputOrderedMapUnstage.Inputs.indicesThe indices inputOrderedMapUnstageNoKey.Inputs.indicesThe indices inputResourceGather.Inputs.indicesThe indices inputResourceGatherNd.Inputs.indicesThe indices inputResourceScatterAdd.Inputs.indicesA tensor of indices into the first dimension ofref.ResourceScatterDiv.Inputs.indicesA tensor of indices into the first dimension ofref.ResourceScatterMax.Inputs.indicesA tensor of indices into the first dimension ofref.ResourceScatterMin.Inputs.indicesA tensor of indices into the first dimension ofref.ResourceScatterMul.Inputs.indicesA tensor of indices into the first dimension ofref.ResourceScatterNdAdd.Inputs.indicesA Tensor.ResourceScatterNdMax.Inputs.indicesA Tensor.ResourceScatterNdMin.Inputs.indicesA Tensor.ResourceScatterNdSub.Inputs.indicesA Tensor.ResourceScatterNdUpdate.Inputs.indicesA Tensor.ResourceScatterSub.Inputs.indicesA tensor of indices into the first dimension ofref.ResourceScatterUpdate.Inputs.indicesA tensor of indices into the first dimension ofref.ScatterAdd.Inputs.indicesA tensor of indices into the first dimension ofref.ScatterDiv.Inputs.indicesA tensor of indices into the first dimension ofref.ScatterMax.Inputs.indicesA tensor of indices into the first dimension ofref.ScatterMin.Inputs.indicesA tensor of indices into the first dimension ofref.ScatterMul.Inputs.indicesA tensor of indices into the first dimension ofref.ScatterNd.Inputs.indicesTensor of indices.ScatterNdAdd.Inputs.indicesA Tensor.ScatterNdMax.Inputs.indicesA Tensor.ScatterNdMin.Inputs.indicesA Tensor.ScatterNdNonAliasingAdd.Inputs.indicesA Tensor.ScatterNdSub.Inputs.indicesA Tensor.ScatterNdUpdate.Inputs.indicesA Tensor.ScatterSub.Inputs.indicesA tensor of indices into the first dimension ofref.ScatterUpdate.Inputs.indicesA tensor of indices into the first dimension ofref.TensorArrayGather.Inputs.indicesThe locations in the TensorArray from which to read tensor elements.TensorArrayScatter.Inputs.indicesThe locations at which to write the tensor elements.TensorListGather.Inputs.indicesThe indices inputTensorListScatter.Inputs.indicesThe indices inputTensorListScatterIntoExistingList.Inputs.indicesThe indices inputTensorScatterNdAdd.Inputs.indicesIndex tensor.TensorScatterNdMax.Inputs.indicesIndex tensor.TensorScatterNdMin.Inputs.indicesIndex tensor.TensorScatterNdSub.Inputs.indicesIndex tensor.TensorScatterNdUpdate.Inputs.indicesIndex tensor.UnravelIndex.Inputs.indicesAn 0-D or 1-DintTensor whose elements are indices into the flattened version of an array of dimensions dims.All.Inputs.inputThe tensor to reduce.Any.Inputs.inputThe tensor to reduce.ApproxTopK.Inputs.inputArray to search.BatchToSpace.Inputs.input4-D tensor with shape[batch*block_size*block_size, height_pad/block_size, width_pad/block_size, depth].BatchToSpaceNd.Inputs.inputN-D with shapeinput_shape = [batch] + spatial_shape + remaining_shape, where spatial_shape has M dimensions.Bitcast.Inputs.inputThe input inputBroadcastTo.Inputs.inputA Tensor to broadcast.Bucketize.Inputs.inputAny shape of Tensor contains with int or float type.Copy.Inputs.inputInput tensor.CopyHost.Inputs.inputInput tensor.CopyToMesh.Inputs.inputThe input inputCopyToMeshGrad.Inputs.inputThe input inputEnsureShape.Inputs.inputA tensor, whose shape is to be validated.ExpandDims.Inputs.inputThe input inputExtractVolumePatches.Inputs.input5-D Tensor with shape[batch, in_planes, in_rows, in_cols, depth].GuaranteeConst.Inputs.inputThe input inputIdentity.Inputs.inputThe input inputKthOrderStatistic.Inputs.inputThe input inputLoopCond.Inputs.inputA boolean scalar, representing the branch predicate of the Switch op.MakeUnique.Inputs.inputThe input inputMax.Inputs.inputThe tensor to reduce.Min.Inputs.inputThe tensor to reduce.MirrorPad.Inputs.inputThe input tensor to be padded.MirrorPadGrad.Inputs.inputThe input tensor to be folded.NcclAllReduce.Inputs.inputThe input inputNcclBroadcast.Inputs.inputThe input inputPad.Inputs.inputThe input inputPlaceholderWithDefault.Inputs.inputThe default value to produce whenoutputis not fed.Print.Inputs.inputThe string scalar to print.Prod.Inputs.inputThe tensor to reduce.Rank.Inputs.inputThe input inputReduceAll.Inputs.inputThe tensor to reduce.ReduceAny.Inputs.inputThe tensor to reduce.ReduceMax.Inputs.inputThe tensor to reduce.ReduceMin.Inputs.inputThe tensor to reduce.ReduceProd.Inputs.inputThe tensor to reduce.ReduceSum.Inputs.inputThe tensor to reduce.RefIdentity.Inputs.inputThe input inputRelayout.Inputs.inputThe input inputRelayoutLike.Inputs.inputThe input inputReverseSequence.Inputs.inputThe input to reverse.Roll.Inputs.inputThe input inputScatterNdNonAliasingAdd.Inputs.inputA Tensor.Shape.Inputs.inputThe input inputSize.Inputs.inputThe input inputSlice.Inputs.inputThe input inputSnapshot.Inputs.inputThe input inputSpaceToBatchNd.Inputs.inputN-D with shapeinput_shape = [batch] + spatial_shape + remaining_shape, where spatial_shape hasMdimensions.Squeeze.Inputs.inputTheinputto squeeze.StochasticCastToInt.Inputs.inputThe operand to stochastically cast to int.StopGradient.Inputs.inputThe input inputStridedSlice.Inputs.inputThe input inputSum.Inputs.inputThe tensor to reduce.TensorStridedSliceUpdate.Inputs.inputThe input inputTile.Inputs.inputCan be of any rank.TopKUnique.Inputs.inputThe input inputTopKWithUnique.Inputs.inputThe input inputVariableShape.Inputs.inputThe input inputTensorListConcatLists.Inputs.inputAThe inputA inputTensorListConcatLists.Inputs.inputBThe inputB inputGetOptions.Inputs.inputDatasetA variant tensor representing the input dataset.TensorListConcat.Inputs.inputHandleThe inputHandle inputTensorListElementShape.Inputs.inputHandleThe inputHandle inputTensorListGather.Inputs.inputHandleThe inputHandle inputTensorListGetItem.Inputs.inputHandleThe inputHandle inputTensorListLength.Inputs.inputHandleThe inputHandle inputTensorListPopBack.Inputs.inputHandleThe inputHandle inputTensorListPushBack.Inputs.inputHandleThe inputHandle inputTensorListResize.Inputs.inputHandleThe inputHandle inputTensorListScatterIntoExistingList.Inputs.inputHandleThe inputHandle inputTensorListSetItem.Inputs.inputHandleThe inputHandle inputTensorListStack.Inputs.inputHandleThe inputHandle inputTensorMapErase.Inputs.inputHandleThe inputHandle inputTensorMapHasKey.Inputs.inputHandleThe inputHandle inputTensorMapInsert.Inputs.inputHandleThe inputHandle inputTensorMapLookup.Inputs.inputHandleThe inputHandle inputTensorMapSize.Inputs.inputHandleThe inputHandle inputTensorMapStackKeys.Inputs.inputHandleThe inputHandle inputTensorListPushBackBatch.Inputs.inputHandlesThe inputHandles inputQuantizedReshape.Inputs.inputMaxThe maximum value of the input.QuantizedReshape.Inputs.inputMinThe minimum value of the input.TensorListSetItem.Inputs.itemThe item inputFileSystemSetConfiguration.Inputs.keyThe name of the configuration option.MapPeek.Inputs.keyThe key inputMapStage.Inputs.keyint64MapUnstage.Inputs.keyThe key inputOrderedMapPeek.Inputs.keyThe key inputOrderedMapStage.Inputs.keyint64OrderedMapUnstage.Inputs.keyThe key inputStochasticCastToInt.Inputs.keyKey for the counter-based RNG algorithm (shape uint64[1]).TensorMapErase.Inputs.keyThe key inputTensorMapHasKey.Inputs.keyThe key inputTensorMapInsert.Inputs.keyThe key inputTensorMapLookup.Inputs.keyThe key inputBarrierInsertMany.Inputs.keysA one-dimensional tensor of keys, with length n.InitializeTable.Inputs.keysKeys of type Tkey.LookupTableFind.Inputs.keysAny shape.LookupTableImport.Inputs.keysAny shape.LookupTableInsert.Inputs.keysAny shape.LookupTableRemove.Inputs.keysAny shape.RelayoutLike.Inputs.layoutInputThe layoutInput inputTensorListConcat.Inputs.leadingDimsThe leadingDims inputTensorArraySplit.Inputs.lengthsThe vector of lengths, how to split the rows of value into the TensorArray.TensorListSplit.Inputs.lengthsThe lengths inputFor.Inputs.limitThe upper bound.Range.Inputs.limit0-D (scalar).UniformQuantizedClipByValue.Inputs.maxThe min value(s) to clip operand.RandomIndexShuffle.Inputs.maxIndexA scalar tensor or vector of dtypedtype.EmptyTensorList.Inputs.maxNumElementsThe maxNumElements inputStackCreate.Inputs.maxSizeThe maximum size of the stack if non-negative.Fingerprint.Inputs.methodFingerprint method used by this op.UniformQuantizedClipByValue.Inputs.minThe min value(s) to clip operand.Tile.Inputs.multiples1-D.MutexLock.Inputs.mutexThe mutex resource to lock.ConsumeMutexLock.Inputs.mutexLockA tensor returned byMutexLock.HistogramFixedWidth.Inputs.nbinsScalarint32 Tensor.LinSpace.Inputs.num0-D tensor.BarrierTakeMany.Inputs.numElementsA single-element tensor containing the number of elements to take.TensorListReserve.Inputs.numElementsThe numElements inputTensorListScatter.Inputs.numElementsThe numElements inputOneHot.Inputs.offValueA scalar defining the value to fill in output whenindices[j] != i.OneHot.Inputs.onValueA scalar defining the value to fill in output whenindices[j] = i.UniformQuantizedClipByValue.Inputs.operandMust be a Tensor of T.UnbatchGrad.Inputs.originalInputThe originalInput inputMirrorPad.Inputs.paddingsA two-column matrix specifying the padding sizes.MirrorPadGrad.Inputs.paddingsA two-column matrix specifying the padding sizes.Pad.Inputs.paddingsThe paddings inputSpaceToBatchNd.Inputs.paddings2-D with shape[M, 2], all values must be >= 0.Gather.Inputs.paramsThe tensor from which to gather values.GatherNd.Inputs.paramsThe tensor from which to gather values.DynamicPartition.Inputs.partitionsAny shape.RefSwitch.Inputs.predA scalar that specifies which output port will receive data.SwitchCond.Inputs.predA scalar that specifies which output port will receive data.Assign.Inputs.refShould be from aVariablenode.AssignAdd.Inputs.refShould be from aVariablenode.AssignSub.Inputs.refShould be from aVariablenode.CountUpTo.Inputs.refShould be from a scalarVariablenode.DestroyTemporaryVariable.Inputs.refA reference to the temporary variable tensor.IsVariableInitialized.Inputs.refShould be from aVariablenode.ResourceScatterNdAdd.Inputs.refA resource handle.ResourceScatterNdMax.Inputs.refA resource handle.ResourceScatterNdMin.Inputs.refA resource handle.ResourceScatterNdSub.Inputs.refA resource handle.ResourceScatterNdUpdate.Inputs.refA resource handle.ResourceStridedSliceAssign.Inputs.refThe ref inputScatterAdd.Inputs.refShould be from aVariablenode.ScatterDiv.Inputs.refShould be from aVariablenode.ScatterMax.Inputs.refShould be from aVariablenode.ScatterMin.Inputs.refShould be from aVariablenode.ScatterMul.Inputs.refShould be from aVariablenode.ScatterNdAdd.Inputs.refA mutable Tensor.ScatterNdMax.Inputs.refA mutable Tensor.ScatterNdMin.Inputs.refA mutable Tensor.ScatterNdSub.Inputs.refA mutable Tensor.ScatterNdUpdate.Inputs.refA mutable Tensor.ScatterSub.Inputs.refShould be from aVariablenode.ScatterUpdate.Inputs.refShould be from aVariablenode.StridedSliceAssign.Inputs.refThe ref inputAssignAddVariableOp.Inputs.resourcehandle to the resource in which to store the variable.AssignSubVariableOp.Inputs.resourcehandle to the resource in which to store the variable.AssignVariableOp.Inputs.resourcehandle to the resource in which to store the variable.DestroyResourceOp.Inputs.resourcehandle to the resource to delete.ReadVariableOp.Inputs.resourcehandle to the resource in which to store the variable.ResourceCountUpTo.Inputs.resourceShould be from a scalarVariablenode.ResourceGather.Inputs.resourceThe resource inputResourceGatherNd.Inputs.resourceThe resource inputResourceScatterAdd.Inputs.resourceShould be from aVariablenode.ResourceScatterDiv.Inputs.resourceShould be from aVariablenode.ResourceScatterMax.Inputs.resourceShould be from aVariablenode.ResourceScatterMin.Inputs.resourceShould be from aVariablenode.ResourceScatterMul.Inputs.resourceShould be from aVariablenode.ResourceScatterSub.Inputs.resourceShould be from aVariablenode.ResourceScatterUpdate.Inputs.resourceShould be from aVariablenode.VarIsInitializedOp.Inputs.resourcethe input resource handle.BroadcastDynamicShape.Inputs.s0The s0 inputBroadcastGradientArgs.Inputs.s0The s0 inputBroadcastDynamicShape.Inputs.s1The s1 inputBroadcastGradientArgs.Inputs.s1The s1 inputUniformQuantizedClipByValue.Inputs.scalesThe float value(s) used as scale(s) when quantizingoperand,minandmax.FileSystemSetConfiguration.Inputs.schemeFile system scheme.RandomIndexShuffle.Inputs.seedA tensor of dtypeTseedand shape [3] or [n, 3].ReverseSequence.Inputs.seqLengths1-D with lengthinput.dims(batch_dim)andmax(seq_lengths) <= input.dims(seq_dim)SetSize.Inputs.setIndices2DTensor, indices of aSparseTensor.SetSize.Inputs.setShape1DTensor, shape of aSparseTensor.SetSize.Inputs.setValues1DTensor, values of aSparseTensor.BroadcastTo.Inputs.shapeAn 1-DintTensor.Empty.Inputs.shape1-D.QuantizedReshape.Inputs.shapeDefines the shape of the output tensor.Reshape.Inputs.shapeDefines the shape of the output tensor.ScatterNd.Inputs.shape1-D.StridedSliceGrad.Inputs.shapeThe shape inputTensorArrayGradWithShape.Inputs.shapeToPrependAn int32 vector representing a shape.Roll.Inputs.shiftDimension must be 0-D or 1-D.Slice.Inputs.sizeOutputsize[i] specifies the number of elements of the 'i'th dimension of 'input' to slice.TensorArray.Inputs.sizeOutputThe size of the array.TensorListResize.Inputs.sizeOutputThe sizeOutput inputEncodeProto.Inputs.sizesTensor of int32 with shape[batch_shape, len(field_names)].SplitV.Inputs.sizeSplitslist containing the sizes of each output tensor along the split dimension.LowerBound.Inputs.sortedInputs2-D Tensor where each row is ordered.UpperBound.Inputs.sortedInputs2-D Tensor where each row is ordered.For.Inputs.startThe lower bound.LinSpace.Inputs.start0-D tensor.Range.Inputs.start0-D (scalar).LinSpace.Inputs.stop0-D tensor.ResourceStridedSliceAssign.Inputs.stridesThe strides inputStridedSlice.Inputs.stridesstrides[i]specifies the increment in theith specification after extracting a given element.StridedSliceAssign.Inputs.stridesThe strides inputStridedSliceGrad.Inputs.stridesThe strides inputTensorStridedSliceUpdate.Inputs.stridesThe strides inputClipByValue.Inputs.tATensor.Select.Inputs.tThe t inputInitializeTable.Inputs.tableHandleHandle to a table which will be initialized.InitializeTableFromTextFile.Inputs.tableHandleHandle to a table which will be initialized.LookupTableExport.Inputs.tableHandleHandle to the table.LookupTableFind.Inputs.tableHandleHandle to the table.LookupTableImport.Inputs.tableHandleHandle to the table.LookupTableInsert.Inputs.tableHandleHandle to the table.LookupTableRemove.Inputs.tableHandleHandle to the table.LookupTableSize.Inputs.tableHandleHandle to the table.RemoteCall.Inputs.targetA fully specified device name where we want to run the function.CheckPinned.Inputs.tensorThe tensor inputQuantizedReshape.Inputs.tensorThe tensor inputReshape.Inputs.tensorThe tensor inputReverse.Inputs.tensorUp to 8-D.Send.Inputs.tensorThe tensor to send.TensorListFromTensor.Inputs.tensorThe tensor inputTensorListPushBack.Inputs.tensorThe tensor inputTensorListPushBackBatch.Inputs.tensorThe tensor inputTensorListScatter.Inputs.tensorThe tensor inputTensorListScatterIntoExistingList.Inputs.tensorThe tensor inputTensorListSplit.Inputs.tensorThe tensor inputTensorScatterNdAdd.Inputs.tensorTensor to copy/update.TensorScatterNdMax.Inputs.tensorTensor to update.TensorScatterNdMin.Inputs.tensorTensor to update.TensorScatterNdSub.Inputs.tensorTensor to copy/update.TensorScatterNdUpdate.Inputs.tensorTensor to copy/update.EditDistance.Inputs.truthIndicesThe indices of the truth list SparseTensor.EditDistance.Inputs.truthShapetruth indices, vector.EditDistance.Inputs.truthValuesThe values of the truth list SparseTensor.ResourceScatterAdd.Inputs.updatesA tensor of updated values to add toref.ResourceScatterDiv.Inputs.updatesA tensor of updated values to add toref.ResourceScatterMax.Inputs.updatesA tensor of updated values to add toref.ResourceScatterMin.Inputs.updatesA tensor of updated values to add toref.ResourceScatterMul.Inputs.updatesA tensor of updated values to add toref.ResourceScatterNdAdd.Inputs.updatesA Tensor.ResourceScatterNdMax.Inputs.updatesA Tensor.ResourceScatterNdMin.Inputs.updatesA Tensor.ResourceScatterNdSub.Inputs.updatesA Tensor.ResourceScatterNdUpdate.Inputs.updatesA Tensor.ResourceScatterSub.Inputs.updatesA tensor of updated values to add toref.ResourceScatterUpdate.Inputs.updatesA tensor of updated values to add toref.ScatterAdd.Inputs.updatesA tensor of updated values to add toref.ScatterDiv.Inputs.updatesA tensor of values thatrefis divided by.ScatterMax.Inputs.updatesA tensor of updated values to reduce intoref.ScatterMin.Inputs.updatesA tensor of updated values to reduce intoref.ScatterMul.Inputs.updatesA tensor of updated values to multiply toref.ScatterNd.Inputs.updatesValues to scatter into the output tensor.ScatterNdAdd.Inputs.updatesA Tensor.ScatterNdMax.Inputs.updatesA Tensor.ScatterNdMin.Inputs.updatesA Tensor.ScatterNdNonAliasingAdd.Inputs.updatesA Tensor.ScatterNdSub.Inputs.updatesA Tensor.ScatterNdUpdate.Inputs.updatesA Tensor.ScatterSub.Inputs.updatesA tensor of updated values to subtract fromref.ScatterUpdate.Inputs.updatesA tensor of updated values to store inref.TensorScatterNdAdd.Inputs.updatesUpdates to scatter into output.TensorScatterNdMax.Inputs.updatesUpdates to scatter into output.TensorScatterNdMin.Inputs.updatesUpdates to scatter into output.TensorScatterNdSub.Inputs.updatesUpdates to scatter into output.TensorScatterNdUpdate.Inputs.updatesUpdates to scatter into output.InplaceAdd.Inputs.vATensorof type T.InplaceSub.Inputs.vATensorof type T.InplaceUpdate.Inputs.vATensorof type T.Assign.Inputs.valueThe value to be assigned to the variable.AssignAdd.Inputs.valueThe value to be added to the variable.AssignAddVariableOp.Inputs.valuethe value by which the variable will be incremented.AssignSub.Inputs.valueThe value to be subtracted to the variable.AssignSubVariableOp.Inputs.valuethe value by which the variable will be incremented.AssignVariableOp.Inputs.valuethe value to set the new tensor to use.FileSystemSetConfiguration.Inputs.valueThe value of the configuration option.Fill.Inputs.value0-D (scalar).GetSessionHandle.Inputs.valueThe tensor to be stored.ResourceStridedSliceAssign.Inputs.valueThe value inputSplit.Inputs.valueThe tensor to split.SplitV.Inputs.valueThe tensor to split.StridedSliceAssign.Inputs.valueThe value inputTensorArrayScatter.Inputs.valueThe concatenated tensor to write to the TensorArray.TensorArraySplit.Inputs.valueThe concatenated tensor to write to the TensorArray.TensorArrayUnpack.Inputs.valueThe value inputTensorArrayWrite.Inputs.valueThe tensor to write to the TensorArray.TensorMapInsert.Inputs.valueThe value inputTensorStridedSliceUpdate.Inputs.valueThe value inputUnstack.Inputs.value1-D or higher, withaxisdimension size equal tonum.HistogramFixedWidth.Inputs.valueRangeShape [2]Tensorof samedtypeasvalues. values <= value_range[0] will be mapped to hist[0], values >= value_range[1] will be mapped to hist[-1].BarrierInsertMany.Inputs.valuesAn any-dimensional tensor of values, which are associated with the respective keys.HistogramFixedWidth.Inputs.valuesNumericTensor.InitializeTable.Inputs.valuesValues of type Tval.LookupTableImport.Inputs.valuesValues to associate with keys.LookupTableInsert.Inputs.valuesValues to associate with keys.LowerBound.Inputs.values2-D Tensor with the same numbers of rows assorted_search_values.UpperBound.Inputs.values2-D Tensor with the same numbers of rows assorted_search_values.DeepCopy.Inputs.xThe source tensor of typeT.InplaceAdd.Inputs.xATensorof type T.InplaceSub.Inputs.xATensorof type T.InplaceUpdate.Inputs.xA tensor of typeT.OnesLike.Inputs.xa tensor of type T.SetDiff1d.Inputs.x1-D.Unique.Inputs.xATensor.UniqueWithCounts.Inputs.xATensor.ZerosLike.Inputs.xa tensor of type T.SetDiff1d.Inputs.y1-D.UniformQuantizedClipByValue.Inputs.zeroPointsThe int32 value(s) used as zero_point(s) when quantizingoperand,minandmax.Fields in org.tensorflow.op.core with type parameters of type OperandModifier and TypeFieldDescriptionPartitionedCall.Inputs.argsA list of input tensors.RemoteCall.Inputs.argsA list of arguments for the function.StatefulPartitionedCall.Inputs.argsA list of input tensors.MapDefun.Inputs.argumentsA list of tensors whose types are `Targuments`, corresponding to the inputs the function should be mapped over.MapDefun.Inputs.capturedInputsA list of tensors whose types are `Tcaptured`, corresponding to the captured inputs of the defun.BatchFunction.Inputs.capturedTensorsThe tensors which are captured in the function, and don't need to be batched.CompositeTensorVariantFromComponents.Inputs.componentsThe component tensors for the extension type value.AssertThat.Inputs.dataThe tensors to print out when condition is false.DynamicStitch.Inputs.dataThe data inputParallelDynamicStitch.Inputs.dataThe data inputDynamicStitch.Inputs.indicesThe indices inputParallelDynamicStitch.Inputs.indicesThe indices inputFor.Inputs.inputA list of input tensors whose types are T.IdentityN.Inputs.inputThe input inputNcclReduce.Inputs.inputThe input inputShapeN.Inputs.inputThe input inputStatefulCase.Inputs.inputA list of input tensors passed to the branch function.StatefulIf.Inputs.inputA list of input tensors.StatefulWhile.Inputs.inputA list of input tensors whose types are T.StatelessCase.Inputs.inputA list of input tensors passed to the branch function.StatelessIf.Inputs.inputA list of input tensors.StatelessWhile.Inputs.inputA list of input tensors whose types are T.Merge.Inputs.inputsThe input tensors, exactly one of which will become available.MlirPassthroughOp.Inputs.inputsThe inputs inputRefMerge.Inputs.inputsThe input tensors, exactly one of which will become available.RefSelect.Inputs.inputsA list of ref tensors, one of which will be forwarded tooutput.Batch.Inputs.inTensorsThe inTensors inputBatchFunction.Inputs.inTensorsThe tensors to be batched.ConcatOffset.Inputs.shapeTheNint32 or int64 vectors representing shape of tensors being concatenated.Concat.Inputs.valuesList ofNTensors to concatenate.EncodeProto.Inputs.valuesList of tensors containing values for the corresponding field.MapStage.Inputs.valuesa list of tensors dtypes A list of data types that inserted values should adhere to.OrderedMapStage.Inputs.valuesa list of tensors dtypes A list of data types that inserted values should adhere to.ParallelConcat.Inputs.valuesTensors to be concatenated.Stack.Inputs.valuesMust be of same shape and type.Stage.Inputs.valuesa list of tensors dtypes A list of data types that inserted values should adhere to.Methods in org.tensorflow.op.core that return OperandModifier and TypeMethodDescriptionCreates a 1-dimensional operand containing the dimensions of a shape followed by the last dimension.Creates a 1-dimensional operand containing the dimensions of a shape followed by the last dimension.Creates a 1-dimensional operand that represents a new shape containing the dimensions of the operand representing a shape, followed by the dimensions of an operand representing a shape to append.static Operand<?> Function.call(Scope scope, ConcreteFunction function, Operand<?> argument) Calls the function in an execution environment, adding its graph as a function if it isn't already present.BooleanMask.create(Scope scope, Operand<T> tensor, Operand<TBool> mask, BooleanMask.Options... options) Apply boolean mask to tensor.BooleanMaskUpdate.create(Scope scope, Operand<T> tensor, Operand<TBool> mask, Operand<T> updates, BooleanMaskUpdate.Options... options) Updates a tensor at the masked values, and returns the updated tensor.Flatten the shape to 1 dimension.Flatten the shape to 1 dimension.Flatten the operand to 1 dimension.Flatten the operand to 1 dimensionCreates a 1-dimensional Operand containing the Shape's first dimension.Creates a 1-dimensional Operand containing the Shape's first dimension.Shapes.numDimensions(Scope scope, Shape<TInt32> shape) Get the number of dimensions of the shape object.Shapes.numDimensions(Scope scope, Shape<U> shape, Class<U> type) Get the number of dimensions of the shape object.Creates a 1-dimensional operand containing the first dimension followed by the dimensions of the shape.Creates a 1-dimensional operand containing the first dimension followed by the dimensions of the shape.Creates a 1-dimensional operand that represents a new shape containing the dimensions of an operand representing the shape to prepend, followed by the dimensions of an operand representing a shape.Reduces the shape to the specified axis.Shapes.reduceDims(Scope scope, Shape<U> shape, Operand<U> axis, Class<U> type) Reduces the shape to the specified axis.Shapes.reduceDims(Scope scope, Operand<T> operand, Operand<TInt32> axis) Reshapes the operand by reducing the shape to the specified axis.Shapes.reduceDims(Scope scope, Operand<T> operand, Operand<U> axis, Class<U> type) Reshapes the operand by reducing the shape to the specified axis.Get the size represented by the TensorFlow shape.Get the size of the specified dimension in the shape.Get the size represented by the TensorFlow shape.Get the size of the specified dimension in the shape.Get the size of the specified dimension for the shape of the tensor.Get the size of the specified dimension for the shape of the tensor.Removes dimensions of size 1 from the shape.Removes dimensions of size 1 from the shape.Creates a 1-dimensional Operand that contains the dimension matching the last dimension of the Shape.Creates a 1-dimensional Operand that contains the dimension matching the last dimension of * the Shape.Creates a 1-dimensional operand with the dimensions matching the first n dimensions of the shape.Creates a 1-dimensional operand containin the dimensions matching the first n dimensions of the shape.Creates a 1-dimensional operand containing the dimensions matching the last n dimensions of the shape.Creates a 1-dimensional operand containing the dimensions matching the last n dimensions of the shape.Methods in org.tensorflow.op.core that return types with arguments of type OperandModifier and TypeMethodDescriptionCalls the function in an execution environment, adding its graph as a function if it isn't already present.BatchFunction.iterator()Case.iterator()CompositeTensorVariantToComponents.iterator()ConcatOffset.iterator()DynamicPartition.iterator()For.iterator()GetElementAtIndex.iterator()Gradients.iterator()IdentityN.iterator()If.iterator()MapDefun.iterator()MapPeek.iterator()MapUnstage.iterator()MlirPassthroughOp.iterator()OrderedMapPeek.iterator()OrderedMapUnstage.iterator()PartitionedCall.iterator()RemoteCall.iterator()ShapeN.iterator()Split.iterator()SplitV.iterator()StagePeek.iterator()StatefulCase.iterator()StatefulIf.iterator()StatefulPartitionedCall.iterator()StatefulWhile.iterator()StatelessCase.iterator()StatelessIf.iterator()StatelessWhile.iterator()Unstack.iterator()Unstage.iterator()While.iterator()Methods in org.tensorflow.op.core with parameters of type OperandModifier and TypeMethodDescriptionCreates a 1-dimensional operand that represents a new shape containing the dimensions of the operand representing a shape, followed by the dimensions of an operand representing a shape to append.static Operand<?> Function.call(Scope scope, ConcreteFunction function, Operand<?> argument) Calls the function in an execution environment, adding its graph as a function if it isn't already present.static AllAll.create(Scope scope, Operand<TBool> input, Operand<? extends TNumber> axis, All.Options... options) Factory method to create a class wrapping a new All operation.static <T extends TType, U extends TType>
AnonymousMutableDenseHashTableAnonymousMutableDenseHashTable.create(Scope scope, Operand<T> emptyKey, Operand<T> deletedKey, Class<U> valueDtype, AnonymousMutableDenseHashTable.Options... options) Factory method to create a class wrapping a new AnonymousMutableDenseHashTable operation.static AnyAny.create(Scope scope, Operand<TBool> input, Operand<? extends TNumber> axis, Any.Options... options) Factory method to create a class wrapping a new Any operation.static <T extends TNumber>
ApproxTopK<T> ApproxTopK.create(Scope scope, Operand<T> input, Long k, ApproxTopK.Options... options) Factory method to create a class wrapping a new ApproxTopK operation.static AssertThatAssertThat.create(Scope scope, Operand<TBool> condition, Iterable<Operand<?>> data, AssertThat.Options... options) Factory method to create a class wrapping a new Assert operation.Assign.create(Scope scope, Operand<T> ref, Operand<T> value, Assign.Options... options) Factory method to create a class wrapping a new Assign operation.AssignAdd.create(Scope scope, Operand<T> ref, Operand<T> value, AssignAdd.Options... options) Factory method to create a class wrapping a new AssignAdd operation.static AssignAddVariableOpAssignAddVariableOp.create(Scope scope, Operand<? extends TType> resource, Operand<? extends TType> value) Factory method to create a class wrapping a new AssignAddVariableOp operation.AssignSub.create(Scope scope, Operand<T> ref, Operand<T> value, AssignSub.Options... options) Factory method to create a class wrapping a new AssignSub operation.static AssignSubVariableOpAssignSubVariableOp.create(Scope scope, Operand<? extends TType> resource, Operand<? extends TType> value) Factory method to create a class wrapping a new AssignSubVariableOp operation.static AssignVariableOpAssignVariableOp.create(Scope scope, Operand<? extends TType> resource, Operand<? extends TType> value, AssignVariableOp.Options... options) Factory method to create a class wrapping a new AssignVariableOp operation.static BarrierCloseBarrierClose.create(Scope scope, Operand<TString> handle, BarrierClose.Options... options) Factory method to create a class wrapping a new BarrierClose operation.static BarrierIncompleteSizeFactory method to create a class wrapping a new BarrierIncompleteSize operation.static BarrierInsertManyBarrierInsertMany.create(Scope scope, Operand<TString> handle, Operand<TString> keys, Operand<? extends TType> values, Long componentIndex) Factory method to create a class wrapping a new BarrierInsertMany operation.static BarrierReadySizeFactory method to create a class wrapping a new BarrierReadySize operation.static BarrierTakeManyBarrierTakeMany.create(Scope scope, Operand<TString> handle, Operand<TInt32> numElements, List<Class<? extends TType>> componentTypes, BarrierTakeMany.Options... options) Factory method to create a class wrapping a new BarrierTakeMany operation.static <T extends TType>
BatchToSpace<T> BatchToSpace.create(Scope scope, Operand<T> input, Operand<? extends TNumber> crops, Long blockSize) Factory method to create a class wrapping a new BatchToSpace operation.static <T extends TType>
BatchToSpaceNd<T> BatchToSpaceNd.create(Scope scope, Operand<T> input, Operand<? extends TNumber> blockShape, Operand<? extends TNumber> crops) Factory method to create a class wrapping a new BatchToSpaceND operation.Factory method to create a class wrapping a new Bitcast operation.BooleanMask.create(Scope scope, Operand<T> tensor, Operand<TBool> mask, BooleanMask.Options... options) Apply boolean mask to tensor.BooleanMaskUpdate.create(Scope scope, Operand<T> tensor, Operand<TBool> mask, Operand<T> updates, BooleanMaskUpdate.Options... options) Updates a tensor at the masked values, and returns the updated tensor.static <T extends TNumber>
BroadcastDynamicShape<T> Factory method to create a class wrapping a new BroadcastArgs operation.static <T extends TNumber>
BroadcastGradientArgs<T> Factory method to create a class wrapping a new BroadcastGradientArgs operation.static <T extends TType>
BroadcastTo<T> Factory method to create a class wrapping a new BroadcastTo operation.static BucketizeFactory method to create a class wrapping a new Bucketize operation.static CaseCase.create(Scope scope, Operand<TInt32> branchIndex, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, List<ConcreteFunction> branches, Case.Options... options) Factory method to create a class wrapping a new Case operation.static <T extends TType>
CheckPinned<T> Factory method to create a class wrapping a new CheckPinned operation.static <T extends TType>
ClipByValue<T> Factory method to create a class wrapping a new ClipByValue operation.CompositeTensorVariantToComponents.create(Scope scope, Operand<? extends TType> encoded, String metadata, List<Class<? extends TType>> Tcomponents) Factory method to create a class wrapping a new CompositeTensorVariantToComponents operation.Factory method to create a class wrapping a new ConcatV2 operation.static <T extends TNumber>
ConcatOffset<T> Factory method to create a class wrapping a new ConcatOffset operation.static ConsumeMutexLockFactory method to create a class wrapping a new ConsumeMutexLock operation.Copy.create(Scope scope, Operand<T> input, Copy.Options... options) Factory method to create a class wrapping a new Copy operation.CopyHost.create(Scope scope, Operand<T> input, CopyHost.Options... options) Factory method to create a class wrapping a new CopyHost operation.static <T extends TType>
CopyToMesh<T> Factory method to create a class wrapping a new CopyToMesh operation.static <T extends TType>
CopyToMeshGrad<T> Factory method to create a class wrapping a new CopyToMeshGrad operation.Factory method to create a class wrapping a new CountUpTo operation.static DecodeProtoDecodeProto.create(Scope scope, Operand<TString> bytes, String messageType, List<String> fieldNames, List<Class<? extends TType>> outputTypes, DecodeProto.Options... options) Factory method to create a class wrapping a new DecodeProtoV2 operation.Factory method to create a class wrapping a new DeepCopy operation.static DeleteSessionTensorFactory method to create a class wrapping a new DeleteSessionTensor operation.static DestroyResourceOpDestroyResourceOp.create(Scope scope, Operand<? extends TType> resource, DestroyResourceOp.Options... options) Factory method to create a class wrapping a new DestroyResourceOp operation.static <T extends TType>
DestroyTemporaryVariable<T> Factory method to create a class wrapping a new DestroyTemporaryVariable operation.static <T extends TType>
DynamicPartition<T> DynamicPartition.create(Scope scope, Operand<T> data, Operand<TInt32> partitions, Long numPartitions) Factory method to create a class wrapping a new DynamicPartition operation.static <T extends TType>
EditDistanceEditDistance.create(Scope scope, Operand<TInt64> hypothesisIndices, Operand<T> hypothesisValues, Operand<TInt64> hypothesisShape, Operand<TInt64> truthIndices, Operand<T> truthValues, Operand<TInt64> truthShape, EditDistance.Options... options) Factory method to create a class wrapping a new EditDistance operation.Factory method to create a class wrapping a new Empty operation.static <U extends TType>
EmptyTensorListEmptyTensorList.create(Scope scope, Operand<? extends TNumber> elementShape, Operand<TInt32> maxNumElements, Class<U> elementDtype) Factory method to create a class wrapping a new EmptyTensorList operation.static EncodeProtoEncodeProto.create(Scope scope, Operand<TInt32> sizes, Iterable<Operand<?>> values, List<String> fieldNames, String messageType, EncodeProto.Options... options) Factory method to create a class wrapping a new EncodeProto operation.static <T extends TType>
EnsureShape<T> Factory method to create a class wrapping a new EnsureShape operation.Enter.create(Scope scope, Operand<T> data, String frameName, Enter.Options... options) Factory method to create a class wrapping a new Enter operation.Factory method to create a class wrapping a new Exit operation.static <T extends TType>
ExpandDims<T> Factory method to create a class wrapping a new ExpandDims operation.static <T extends TNumber>
ExtractVolumePatches<T> ExtractVolumePatches.create(Scope scope, Operand<T> input, List<Long> ksizes, List<Long> strides, String padding) Factory method to create a class wrapping a new ExtractVolumePatches operation.static FileSystemSetConfigurationFileSystemSetConfiguration.create(Scope scope, Operand<TString> scheme, Operand<TString> key, Operand<TString> value) Factory method to create a class wrapping a new FileSystemSetConfiguration operation.Factory method to create a class wrapping a new Fill operation.static FingerprintFactory method to create a class wrapping a new Fingerprint operation.static ForFor.create(Scope scope, Operand<TInt32> start, Operand<TInt32> limit, Operand<TInt32> delta, Iterable<Operand<?>> input, ConcreteFunction body) Factory method to create a class wrapping a new For operation.Gather.create(Scope scope, Operand<T> params, Operand<? extends TNumber> indices, Operand<? extends TNumber> axis, Gather.Options... options) Factory method to create a class wrapping a new GatherV2 operation.GatherNd.create(Scope scope, Operand<T> params, Operand<? extends TNumber> indices, GatherNd.Options... options) Factory method to create a class wrapping a new GatherNd operation.static GetElementAtIndexGetElementAtIndex.create(Scope scope, Operand<? extends TType> dataset, Operand<TInt64> index, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new GetElementAtIndex operation.static GetOptionsFactory method to create a class wrapping a new GetOptions operation.static GetSessionHandleFactory method to create a class wrapping a new GetSessionHandleV2 operation.static <T extends TType>
GetSessionTensor<T> Factory method to create a class wrapping a new GetSessionTensor operation.static GradientsGradients.create(Scope scope, Operand<?> y, Iterable<? extends Operand<?>> x, Gradients.Options... options) Adds gradients computation ops to the graph according to scope.static <T extends TType>
GuaranteeConst<T> Factory method to create a class wrapping a new GuaranteeConst operation.static <T extends TNumber>
HistogramFixedWidth<TInt32> HistogramFixedWidth.create(Scope scope, Operand<T> values, Operand<T> valueRange, Operand<TInt32> nbins) Factory method to create a class wrapping a new HistogramFixedWidth operation, with the default output types.static <U extends TNumber, T extends TNumber>
HistogramFixedWidth<U> HistogramFixedWidth.create(Scope scope, Operand<T> values, Operand<T> valueRange, Operand<TInt32> nbins, Class<U> dtype) Factory method to create a class wrapping a new HistogramFixedWidth operation.Factory method to create a class wrapping a new Identity operation.static IfIf.create(Scope scope, Operand<? extends TType> cond, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, ConcreteFunction thenBranch, ConcreteFunction elseBranch, If.Options... options) Factory method to create a class wrapping a new If operation.static InitializeTableInitializeTable.create(Scope scope, Operand<? extends TType> tableHandle, Operand<? extends TType> keys, Operand<? extends TType> values) Factory method to create a class wrapping a new InitializeTableV2 operation.static InitializeTableFromTextFileInitializeTableFromTextFile.create(Scope scope, Operand<? extends TType> tableHandle, Operand<TString> filename, Long keyIndex, Long valueIndex, InitializeTableFromTextFile.Options... options) Factory method to create a class wrapping a new InitializeTableFromTextFileV2 operation.static <T extends TType>
InplaceAdd<T> Factory method to create a class wrapping a new InplaceAdd operation.static <T extends TType>
InplaceSub<T> Factory method to create a class wrapping a new InplaceSub operation.static <T extends TType>
InplaceUpdate<T> Factory method to create a class wrapping a new InplaceUpdate operation.static IsVariableInitializedFactory method to create a class wrapping a new IsVariableInitialized operation.static KthOrderStatisticFactory method to create a class wrapping a new KthOrderStatistic operation.Factory method to create a class wrapping a new LinSpace operation.static <T extends TType, U extends TType>
LookupTableExport<T, U> LookupTableExport.create(Scope scope, Operand<? extends TType> tableHandle, Class<T> Tkeys, Class<U> Tvalues) Factory method to create a class wrapping a new LookupTableExportV2 operation.static <U extends TType>
LookupTableFind<U> LookupTableFind.create(Scope scope, Operand<? extends TType> tableHandle, Operand<? extends TType> keys, Operand<U> defaultValue) Factory method to create a class wrapping a new LookupTableFindV2 operation.static LookupTableImportLookupTableImport.create(Scope scope, Operand<? extends TType> tableHandle, Operand<? extends TType> keys, Operand<? extends TType> values) Factory method to create a class wrapping a new LookupTableImportV2 operation.static LookupTableInsertLookupTableInsert.create(Scope scope, Operand<? extends TType> tableHandle, Operand<? extends TType> keys, Operand<? extends TType> values) Factory method to create a class wrapping a new LookupTableInsertV2 operation.static LookupTableRemoveLookupTableRemove.create(Scope scope, Operand<? extends TType> tableHandle, Operand<? extends TType> keys) Factory method to create a class wrapping a new LookupTableRemoveV2 operation.static LookupTableSizeFactory method to create a class wrapping a new LookupTableSizeV2 operation.static LoopCondFactory method to create a class wrapping a new LoopCond operation.static <T extends TType>
LowerBound<TInt32> Factory method to create a class wrapping a new LowerBound operation, with the default output types.static <U extends TNumber, T extends TType>
LowerBound<U> Factory method to create a class wrapping a new LowerBound operation.static MakeUniqueFactory method to create a class wrapping a new MakeUnique operation.static MapPeekMapPeek.create(Scope scope, Operand<TInt64> key, Operand<TInt32> indices, List<Class<? extends TType>> dtypes, MapPeek.Options... options) Factory method to create a class wrapping a new MapPeek operation.static MapStageMapStage.create(Scope scope, Operand<TInt64> key, Operand<TInt32> indices, Iterable<Operand<?>> values, List<Class<? extends TType>> dtypes, MapStage.Options... options) Factory method to create a class wrapping a new MapStage operation.static MapUnstageMapUnstage.create(Scope scope, Operand<TInt64> key, Operand<TInt32> indices, List<Class<? extends TType>> dtypes, MapUnstage.Options... options) Factory method to create a class wrapping a new MapUnstage operation.static MapUnstageNoKeyMapUnstageNoKey.create(Scope scope, Operand<TInt32> indices, List<Class<? extends TType>> dtypes, MapUnstageNoKey.Options... options) Factory method to create a class wrapping a new MapUnstageNoKey operation.Factory method to create a class wrapping a new Max operation.Factory method to create a class wrapping a new Min operation.Factory method to create a class wrapping a new MirrorPad operation.static <T extends TType>
MirrorPadGrad<T> MirrorPadGrad.create(Scope scope, Operand<T> input, Operand<? extends TNumber> paddings, String mode) Factory method to create a class wrapping a new MirrorPadGrad operation.static <T extends TType, U extends TType>
MutableDenseHashTableMutableDenseHashTable.create(Scope scope, Operand<T> emptyKey, Operand<T> deletedKey, Class<U> valueDtype, MutableDenseHashTable.Options... options) Factory method to create a class wrapping a new MutableDenseHashTableV2 operation.static MutexLockFactory method to create a class wrapping a new MutexLock operation.static <T extends TNumber>
NcclAllReduce<T> NcclAllReduce.create(Scope scope, Operand<T> input, String reduction, Long numDevices, String sharedName) Deprecated.Factory method to create a class wrapping a new NcclAllReduce operation.static <T extends TNumber>
NcclBroadcast<T> Deprecated.Factory method to create a class wrapping a new NcclBroadcast operation.static <T extends TType>
NextIteration<T> Factory method to create a class wrapping a new NextIteration operation.OneHot.create(Scope scope, Operand<? extends TNumber> indices, Operand<TInt32> depth, Operand<U> onValue, Operand<U> offValue, OneHot.Options... options) Factory method to create a class wrapping a new OneHot operation.Creates a one valued tensor given its type and shape.Factory method to create a class wrapping a new OnesLike operation.static OrderedMapPeekOrderedMapPeek.create(Scope scope, Operand<TInt64> key, Operand<TInt32> indices, List<Class<? extends TType>> dtypes, OrderedMapPeek.Options... options) Factory method to create a class wrapping a new OrderedMapPeek operation.static OrderedMapStageOrderedMapStage.create(Scope scope, Operand<TInt64> key, Operand<TInt32> indices, Iterable<Operand<?>> values, List<Class<? extends TType>> dtypes, OrderedMapStage.Options... options) Factory method to create a class wrapping a new OrderedMapStage operation.static OrderedMapUnstageOrderedMapUnstage.create(Scope scope, Operand<TInt64> key, Operand<TInt32> indices, List<Class<? extends TType>> dtypes, OrderedMapUnstage.Options... options) Factory method to create a class wrapping a new OrderedMapUnstage operation.static OrderedMapUnstageNoKeyOrderedMapUnstageNoKey.create(Scope scope, Operand<TInt32> indices, List<Class<? extends TType>> dtypes, OrderedMapUnstageNoKey.Options... options) Factory method to create a class wrapping a new OrderedMapUnstageNoKey operation.Pad.create(Scope scope, Operand<T> input, Operand<? extends TNumber> paddings, Operand<T> constantValues) Factory method to create a class wrapping a new PadV2 operation.static <T extends TType>
PlaceholderWithDefault<T> Factory method to create a class wrapping a new PlaceholderWithDefault operation.static PrintPrint.create(Scope scope, Operand<TString> input, Print.Options... options) Factory method to create a class wrapping a new PrintV2 operation.Prod.create(Scope scope, Operand<T> input, Operand<? extends TNumber> axis, Prod.Options... options) Factory method to create a class wrapping a new Prod operation.static <T extends TType>
QuantizedReshape<T> QuantizedReshape.create(Scope scope, Operand<T> tensor, Operand<? extends TNumber> shape, Operand<TFloat32> inputMin, Operand<TFloat32> inputMax) Factory method to create a class wrapping a new QuantizedReshape operation.static <T extends TNumber>
RandomIndexShuffle<T> RandomIndexShuffle.create(Scope scope, Operand<T> index, Operand<? extends TNumber> seed, Operand<T> maxIndex, RandomIndexShuffle.Options... options) Factory method to create a class wrapping a new RandomIndexShuffle operation.Factory method to create a class wrapping a new Range operation.static RankFactory method to create a class wrapping a new Rank operation.static <T extends TType>
ReadVariableOp<T> Factory method to create a class wrapping a new ReadVariableOp operation.static ReduceAllReduceAll.create(Scope scope, Operand<TBool> input, Operand<? extends TNumber> axis, ReduceAll.Options... options) Factory method to create a class wrapping a new All operation.static ReduceAnyReduceAny.create(Scope scope, Operand<TBool> input, Operand<? extends TNumber> axis, ReduceAny.Options... options) Factory method to create a class wrapping a new Any operation.ReduceMax.create(Scope scope, Operand<T> input, Operand<? extends TNumber> axis, ReduceMax.Options... options) Factory method to create a class wrapping a new Max operation.ReduceMin.create(Scope scope, Operand<T> input, Operand<? extends TNumber> axis, ReduceMin.Options... options) Factory method to create a class wrapping a new Min operation.static <T extends TType>
ReduceProd<T> ReduceProd.create(Scope scope, Operand<T> input, Operand<? extends TNumber> axis, ReduceProd.Options... options) Factory method to create a class wrapping a new Prod operation.ReduceSum.create(Scope scope, Operand<T> input, Operand<? extends TNumber> axis, ReduceSum.Options... options) Factory method to create a class wrapping a new Sum operation.RefEnter.create(Scope scope, Operand<T> data, String frameName, RefEnter.Options... options) Factory method to create a class wrapping a new RefEnter operation.Factory method to create a class wrapping a new RefExit operation.static <T extends TType>
RefIdentity<T> Factory method to create a class wrapping a new RefIdentity operation.static <T extends TType>
RefNextIteration<T> Factory method to create a class wrapping a new RefNextIteration operation.Factory method to create a class wrapping a new RefSelect operation.Factory method to create a class wrapping a new RefSwitch operation.Factory method to create a class wrapping a new Relayout operation.static <T extends TType>
RelayoutLike<T> Factory method to create a class wrapping a new RelayoutLike operation.static RemoteCallRemoteCall.create(Scope scope, Operand<TString> target, Iterable<Operand<?>> args, List<Class<? extends TType>> Tout, ConcreteFunction f) Factory method to create a class wrapping a new RemoteCall operation.Factory method to create a class wrapping a new Reshape operation.static <T extends TNumber>
ResourceCountUpTo<T> Factory method to create a class wrapping a new ResourceCountUpTo operation.static <U extends TType>
ResourceGather<U> ResourceGather.create(Scope scope, Operand<? extends TType> resource, Operand<? extends TNumber> indices, Class<U> dtype, ResourceGather.Options... options) Factory method to create a class wrapping a new ResourceGather operation.static <U extends TType>
ResourceGatherNd<U> ResourceGatherNd.create(Scope scope, Operand<? extends TType> resource, Operand<? extends TNumber> indices, Class<U> dtype) Factory method to create a class wrapping a new ResourceGatherNd operation.static ResourceScatterAddResourceScatterAdd.create(Scope scope, Operand<? extends TType> resource, Operand<? extends TNumber> indices, Operand<? extends TType> updates) Factory method to create a class wrapping a new ResourceScatterAdd operation.static ResourceScatterDivResourceScatterDiv.create(Scope scope, Operand<? extends TType> resource, Operand<? extends TNumber> indices, Operand<? extends TType> updates) Factory method to create a class wrapping a new ResourceScatterDiv operation.static ResourceScatterMaxResourceScatterMax.create(Scope scope, Operand<? extends TType> resource, Operand<? extends TNumber> indices, Operand<? extends TType> updates) Factory method to create a class wrapping a new ResourceScatterMax operation.static ResourceScatterMinResourceScatterMin.create(Scope scope, Operand<? extends TType> resource, Operand<? extends TNumber> indices, Operand<? extends TType> updates) Factory method to create a class wrapping a new ResourceScatterMin operation.static ResourceScatterMulResourceScatterMul.create(Scope scope, Operand<? extends TType> resource, Operand<? extends TNumber> indices, Operand<? extends TType> updates) Factory method to create a class wrapping a new ResourceScatterMul operation.static ResourceScatterNdAddResourceScatterNdAdd.create(Scope scope, Operand<? extends TType> ref, Operand<? extends TNumber> indices, Operand<? extends TType> updates, ResourceScatterNdAdd.Options... options) Factory method to create a class wrapping a new ResourceScatterNdAdd operation.static ResourceScatterNdMaxResourceScatterNdMax.create(Scope scope, Operand<? extends TType> ref, Operand<? extends TNumber> indices, Operand<? extends TType> updates, ResourceScatterNdMax.Options... options) Factory method to create a class wrapping a new ResourceScatterNdMax operation.static ResourceScatterNdMinResourceScatterNdMin.create(Scope scope, Operand<? extends TType> ref, Operand<? extends TNumber> indices, Operand<? extends TType> updates, ResourceScatterNdMin.Options... options) Factory method to create a class wrapping a new ResourceScatterNdMin operation.static ResourceScatterNdSubResourceScatterNdSub.create(Scope scope, Operand<? extends TType> ref, Operand<? extends TNumber> indices, Operand<? extends TType> updates, ResourceScatterNdSub.Options... options) Factory method to create a class wrapping a new ResourceScatterNdSub operation.static ResourceScatterNdUpdateResourceScatterNdUpdate.create(Scope scope, Operand<? extends TType> ref, Operand<? extends TNumber> indices, Operand<? extends TType> updates, ResourceScatterNdUpdate.Options... options) Factory method to create a class wrapping a new ResourceScatterNdUpdate operation.static ResourceScatterSubResourceScatterSub.create(Scope scope, Operand<? extends TType> resource, Operand<? extends TNumber> indices, Operand<? extends TType> updates) Factory method to create a class wrapping a new ResourceScatterSub operation.static ResourceScatterUpdateResourceScatterUpdate.create(Scope scope, Operand<? extends TType> resource, Operand<? extends TNumber> indices, Operand<? extends TType> updates) Factory method to create a class wrapping a new ResourceScatterUpdate operation.static <T extends TNumber>
ResourceStridedSliceAssignResourceStridedSliceAssign.create(Scope scope, Operand<? extends TType> ref, Operand<T> begin, Operand<T> end, Operand<T> strides, Operand<? extends TType> value, ResourceStridedSliceAssign.Options... options) Factory method to create a class wrapping a new ResourceStridedSliceAssign operation.Factory method to create a class wrapping a new ReverseV2 operation.static <T extends TType>
ReverseSequence<T> ReverseSequence.create(Scope scope, Operand<T> input, Operand<? extends TNumber> seqLengths, Long seqDim, ReverseSequence.Options... options) Factory method to create a class wrapping a new ReverseSequence operation.Roll.create(Scope scope, Operand<T> input, Operand<? extends TNumber> shift, Operand<? extends TNumber> axis) Factory method to create a class wrapping a new Roll operation.static <T extends TType>
ScatterAdd<T> ScatterAdd.create(Scope scope, Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterAdd.Options... options) Factory method to create a class wrapping a new ScatterAdd operation.static <T extends TType>
ScatterDiv<T> ScatterDiv.create(Scope scope, Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterDiv.Options... options) Factory method to create a class wrapping a new ScatterDiv operation.static <T extends TNumber>
ScatterMax<T> ScatterMax.create(Scope scope, Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterMax.Options... options) Factory method to create a class wrapping a new ScatterMax operation.static <T extends TNumber>
ScatterMin<T> ScatterMin.create(Scope scope, Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterMin.Options... options) Factory method to create a class wrapping a new ScatterMin operation.static <T extends TType>
ScatterMul<T> ScatterMul.create(Scope scope, Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterMul.Options... options) Factory method to create a class wrapping a new ScatterMul operation.ScatterNd.create(Scope scope, Operand<T> indices, Operand<U> updates, Operand<T> shape, ScatterNd.Options... options) Factory method to create a class wrapping a new ScatterNd operation.static <T extends TType>
ScatterNdAdd<T> ScatterNdAdd.create(Scope scope, Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterNdAdd.Options... options) Factory method to create a class wrapping a new ScatterNdAdd operation.static <T extends TType>
ScatterNdMax<T> ScatterNdMax.create(Scope scope, Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterNdMax.Options... options) Factory method to create a class wrapping a new ScatterNdMax operation.static <T extends TType>
ScatterNdMin<T> ScatterNdMin.create(Scope scope, Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterNdMin.Options... options) Factory method to create a class wrapping a new ScatterNdMin operation.static <T extends TType>
ScatterNdNonAliasingAdd<T> ScatterNdNonAliasingAdd.create(Scope scope, Operand<T> input, Operand<? extends TNumber> indices, Operand<T> updates, ScatterNdNonAliasingAdd.Options... options) Factory method to create a class wrapping a new ScatterNdNonAliasingAdd operation.static <T extends TType>
ScatterNdSub<T> ScatterNdSub.create(Scope scope, Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterNdSub.Options... options) Factory method to create a class wrapping a new ScatterNdSub operation.static <T extends TType>
ScatterNdUpdate<T> ScatterNdUpdate.create(Scope scope, Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterNdUpdate.Options... options) Factory method to create a class wrapping a new ScatterNdUpdate operation.static <T extends TType>
ScatterSub<T> ScatterSub.create(Scope scope, Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterSub.Options... options) Factory method to create a class wrapping a new ScatterSub operation.static <T extends TType>
ScatterUpdate<T> ScatterUpdate.create(Scope scope, Operand<T> ref, Operand<? extends TNumber> indices, Operand<T> updates, ScatterUpdate.Options... options) Factory method to create a class wrapping a new ScatterUpdate operation.Factory method to create a class wrapping a new SelectV2 operation.static SendSend.create(Scope scope, Operand<? extends TType> tensor, String tensorName, String sendDevice, Long sendDeviceIncarnation, String recvDevice, Send.Options... options) Factory method to create a class wrapping a new Send operation.Factory method to create a class wrapping a new ListDiff operation, with the default output types.Factory method to create a class wrapping a new ListDiff operation.static SetSizeSetSize.create(Scope scope, Operand<TInt64> setIndices, Operand<? extends TType> setValues, Operand<TInt64> setShape, SetSize.Options... options) Factory method to create a class wrapping a new SetSize operation.Factory method to create a class wrapping a new Shape operation, with the default output types.Factory method to create a class wrapping a new Shape operation.Factory method to create a class wrapping a new Size operation, with the default output types.Factory method to create a class wrapping a new Size operation.Factory method to create a class wrapping a new Slice operation.Factory method to create a class wrapping a new Snapshot operation.static <T extends TType>
SpaceToBatchNd<T> SpaceToBatchNd.create(Scope scope, Operand<T> input, Operand<? extends TNumber> blockShape, Operand<? extends TNumber> paddings) Factory method to create a class wrapping a new SpaceToBatchND operation.Factory method to create a class wrapping a new Split operation.SplitV.create(Scope scope, Operand<T> value, Operand<? extends TNumber> sizeSplits, Operand<TInt32> axis, Long numSplit) Factory method to create a class wrapping a new SplitV operation.Squeeze.create(Scope scope, Operand<T> input, Squeeze.Options... options) Factory method to create a class wrapping a new Squeeze operation.static StackCloseFactory method to create a class wrapping a new StackCloseV2 operation.static <T extends TType>
StackCreateStackCreate.create(Scope scope, Operand<TInt32> maxSize, Class<T> elemType, StackCreate.Options... options) Factory method to create a class wrapping a new StackV2 operation.Factory method to create a class wrapping a new StackPopV2 operation.StackPush.create(Scope scope, Operand<? extends TType> handle, Operand<T> elem, StackPush.Options... options) Factory method to create a class wrapping a new StackPushV2 operation.static StagePeekStagePeek.create(Scope scope, Operand<TInt32> index, List<Class<? extends TType>> dtypes, StagePeek.Options... options) Factory method to create a class wrapping a new StagePeek operation.static StatefulCaseStatefulCase.create(Scope scope, Operand<TInt32> branchIndex, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, List<ConcreteFunction> branches, Case.Options... options) Factory method to create a class wrapping a new Case operation.static StatefulIfStatefulIf.create(Scope scope, Operand<? extends TType> cond, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, ConcreteFunction thenBranch, ConcreteFunction elseBranch, If.Options... options) Factory method to create a class wrapping a new If operation.static StatelessCaseStatelessCase.create(Scope scope, Operand<TInt32> branchIndex, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, List<ConcreteFunction> branches, Case.Options... options) Factory method to create a class wrapping a new StatelessCase operation.static StatelessIfStatelessIf.create(Scope scope, Operand<? extends TType> cond, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, ConcreteFunction thenBranch, ConcreteFunction elseBranch, If.Options... options) Factory method to create a class wrapping a new StatelessIf operation.static <U extends TNumber>
StochasticCastToInt<U> StochasticCastToInt.create(Scope scope, Operand<? extends TNumber> input, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg, Class<U> Tout) Factory method to create a class wrapping a new StochasticCastToInt operation.static <T extends TType>
StopGradient<T> Factory method to create a class wrapping a new StopGradient operation.static <T extends TType, U extends TNumber>
StridedSlice<T> StridedSlice.create(Scope scope, Operand<T> input, Operand<U> begin, Operand<U> end, Operand<U> strides, StridedSlice.Options... options) Factory method to create a class wrapping a new StridedSlice operation.static <T extends TType, U extends TNumber>
StridedSliceAssign<T> StridedSliceAssign.create(Scope scope, Operand<T> ref, Operand<U> begin, Operand<U> end, Operand<U> strides, Operand<T> value, StridedSliceAssign.Options... options) Factory method to create a class wrapping a new StridedSliceAssign operation.static <U extends TType, T extends TNumber>
StridedSliceGrad<U> StridedSliceGrad.create(Scope scope, Operand<T> shape, Operand<T> begin, Operand<T> end, Operand<T> strides, Operand<U> dy, StridedSliceGrad.Options... options) Factory method to create a class wrapping a new StridedSliceGrad operation.Factory method to create a class wrapping a new Sum operation.static <T extends TType>
SwitchCond<T> Factory method to create a class wrapping a new Switch operation.static <T extends TType>
TensorArrayTensorArray.create(Scope scope, Operand<TInt32> sizeOutput, Class<T> dtype, TensorArray.Options... options) Factory method to create a class wrapping a new TensorArrayV3 operation.static TensorArrayCloseFactory method to create a class wrapping a new TensorArrayCloseV3 operation.static <T extends TType>
TensorArrayConcat<T> TensorArrayConcat.create(Scope scope, Operand<? extends TType> handle, Operand<TFloat32> flowIn, Class<T> dtype, TensorArrayConcat.Options... options) Factory method to create a class wrapping a new TensorArrayConcatV3 operation.static <T extends TType>
TensorArrayGather<T> TensorArrayGather.create(Scope scope, Operand<? extends TType> handle, Operand<TInt32> indices, Operand<TFloat32> flowIn, Class<T> dtype, TensorArrayGather.Options... options) Factory method to create a class wrapping a new TensorArrayGatherV3 operation.static TensorArrayGradTensorArrayGrad.create(Scope scope, Operand<? extends TType> handle, Operand<TFloat32> flowIn, String source) Factory method to create a class wrapping a new TensorArrayGradV3 operation.static TensorArrayGradWithShapeTensorArrayGradWithShape.create(Scope scope, Operand<? extends TType> handle, Operand<TFloat32> flowIn, Operand<TInt32> shapeToPrepend, String source) Factory method to create a class wrapping a new TensorArrayGradWithShape operation.static <T extends TType>
TensorArrayPack<T> TensorArrayPack.create(Scope scope, Operand<TString> handle, Operand<TFloat32> flowIn, Class<T> dtype, TensorArrayPack.Options... options) Factory method to create a class wrapping a new TensorArrayPack operation.static <T extends TType>
TensorArrayRead<T> TensorArrayRead.create(Scope scope, Operand<? extends TType> handle, Operand<TInt32> index, Operand<TFloat32> flowIn, Class<T> dtype) Factory method to create a class wrapping a new TensorArrayReadV3 operation.static TensorArrayScatterTensorArrayScatter.create(Scope scope, Operand<? extends TType> handle, Operand<TInt32> indices, Operand<? extends TType> value, Operand<TFloat32> flowIn) Factory method to create a class wrapping a new TensorArrayScatterV3 operation.static TensorArraySizeFactory method to create a class wrapping a new TensorArraySizeV3 operation.static TensorArraySplitTensorArraySplit.create(Scope scope, Operand<? extends TType> handle, Operand<? extends TType> value, Operand<TInt64> lengths, Operand<TFloat32> flowIn) Factory method to create a class wrapping a new TensorArraySplitV3 operation.static TensorArrayUnpackTensorArrayUnpack.create(Scope scope, Operand<TString> handle, Operand<? extends TType> value, Operand<TFloat32> flowIn) Factory method to create a class wrapping a new TensorArrayUnpack operation.static TensorArrayWriteTensorArrayWrite.create(Scope scope, Operand<? extends TType> handle, Operand<TInt32> index, Operand<? extends TType> value, Operand<TFloat32> flowIn) Factory method to create a class wrapping a new TensorArrayWriteV3 operation.static <U extends TType>
TensorListConcat<U> TensorListConcat.create(Scope scope, Operand<? extends TType> inputHandle, Operand<? extends TNumber> elementShape, Operand<TInt64> leadingDims, Class<U> elementDtype) Factory method to create a class wrapping a new TensorListConcatV2 operation.static <T extends TType>
TensorListConcatListsTensorListConcatLists.create(Scope scope, Operand<? extends TType> inputA, Operand<? extends TType> inputB, Class<T> elementDtype) Factory method to create a class wrapping a new TensorListConcatLists operation.static <T extends TNumber>
TensorListElementShape<T> TensorListElementShape.create(Scope scope, Operand<? extends TType> inputHandle, Class<T> shapeType) Factory method to create a class wrapping a new TensorListElementShape operation.static TensorListFromTensorTensorListFromTensor.create(Scope scope, Operand<? extends TType> tensor, Operand<? extends TNumber> elementShape) Factory method to create a class wrapping a new TensorListFromTensor operation.static <T extends TType>
TensorListGather<T> TensorListGather.create(Scope scope, Operand<? extends TType> inputHandle, Operand<TInt32> indices, Operand<TInt32> elementShape, Class<T> elementDtype) Factory method to create a class wrapping a new TensorListGather operation.static <T extends TType>
TensorListGetItem<T> TensorListGetItem.create(Scope scope, Operand<? extends TType> inputHandle, Operand<TInt32> index, Operand<TInt32> elementShape, Class<T> elementDtype) Factory method to create a class wrapping a new TensorListGetItem operation.static TensorListLengthFactory method to create a class wrapping a new TensorListLength operation.static <T extends TType>
TensorListPopBack<T> TensorListPopBack.create(Scope scope, Operand<? extends TType> inputHandle, Operand<TInt32> elementShape, Class<T> elementDtype) Factory method to create a class wrapping a new TensorListPopBack operation.static TensorListPushBackTensorListPushBack.create(Scope scope, Operand<? extends TType> inputHandle, Operand<? extends TType> tensor) Factory method to create a class wrapping a new TensorListPushBack operation.static TensorListPushBackBatchTensorListPushBackBatch.create(Scope scope, Operand<? extends TType> inputHandles, Operand<? extends TType> tensor) Factory method to create a class wrapping a new TensorListPushBackBatch operation.static <U extends TType>
TensorListReserveTensorListReserve.create(Scope scope, Operand<? extends TNumber> elementShape, Operand<TInt32> numElements, Class<U> elementDtype) Factory method to create a class wrapping a new TensorListReserve operation.static TensorListResizeTensorListResize.create(Scope scope, Operand<? extends TType> inputHandle, Operand<TInt32> sizeOutput) Factory method to create a class wrapping a new TensorListResize operation.static TensorListScatterTensorListScatter.create(Scope scope, Operand<? extends TType> tensor, Operand<TInt32> indices, Operand<? extends TNumber> elementShape, Operand<TInt32> numElements) Factory method to create a class wrapping a new TensorListScatterV2 operation.TensorListScatterIntoExistingList.create(Scope scope, Operand<? extends TType> inputHandle, Operand<? extends TType> tensor, Operand<TInt32> indices) Factory method to create a class wrapping a new TensorListScatterIntoExistingList operation.static TensorListSetItemTensorListSetItem.create(Scope scope, Operand<? extends TType> inputHandle, Operand<TInt32> index, Operand<? extends TType> item, TensorListSetItem.Options... options) Factory method to create a class wrapping a new TensorListSetItem operation.static TensorListSplitTensorListSplit.create(Scope scope, Operand<? extends TType> tensor, Operand<? extends TNumber> elementShape, Operand<TInt64> lengths) Factory method to create a class wrapping a new TensorListSplit operation.static <T extends TType>
TensorListStack<T> TensorListStack.create(Scope scope, Operand<? extends TType> inputHandle, Operand<TInt32> elementShape, Class<T> elementDtype, TensorListStack.Options... options) Factory method to create a class wrapping a new TensorListStack operation.static <U extends TType>
TensorMapEraseTensorMapErase.create(Scope scope, Operand<? extends TType> inputHandle, Operand<? extends TType> key, Class<U> valueDtype) Factory method to create a class wrapping a new TensorMapErase operation.static TensorMapHasKeyTensorMapHasKey.create(Scope scope, Operand<? extends TType> inputHandle, Operand<? extends TType> key) Factory method to create a class wrapping a new TensorMapHasKey operation.static TensorMapInsertTensorMapInsert.create(Scope scope, Operand<? extends TType> inputHandle, Operand<? extends TType> key, Operand<? extends TType> value) Factory method to create a class wrapping a new TensorMapInsert operation.static <U extends TType>
TensorMapLookup<U> TensorMapLookup.create(Scope scope, Operand<? extends TType> inputHandle, Operand<? extends TType> key, Class<U> valueDtype) Factory method to create a class wrapping a new TensorMapLookup operation.static TensorMapSizeFactory method to create a class wrapping a new TensorMapSize operation.static <T extends TType>
TensorMapStackKeys<T> Factory method to create a class wrapping a new TensorMapStackKeys operation.static <T extends TType>
TensorScatterNdAdd<T> TensorScatterNdAdd.create(Scope scope, Operand<T> tensor, Operand<? extends TNumber> indices, Operand<T> updates, TensorScatterNdAdd.Options... options) Factory method to create a class wrapping a new TensorScatterAdd operation.static <T extends TType>
TensorScatterNdMax<T> TensorScatterNdMax.create(Scope scope, Operand<T> tensor, Operand<? extends TNumber> indices, Operand<T> updates, TensorScatterNdMax.Options... options) Factory method to create a class wrapping a new TensorScatterMax operation.static <T extends TType>
TensorScatterNdMin<T> TensorScatterNdMin.create(Scope scope, Operand<T> tensor, Operand<? extends TNumber> indices, Operand<T> updates, TensorScatterNdMin.Options... options) Factory method to create a class wrapping a new TensorScatterMin operation.static <T extends TType>
TensorScatterNdSub<T> TensorScatterNdSub.create(Scope scope, Operand<T> tensor, Operand<? extends TNumber> indices, Operand<T> updates, TensorScatterNdSub.Options... options) Factory method to create a class wrapping a new TensorScatterSub operation.static <T extends TType>
TensorScatterNdUpdate<T> TensorScatterNdUpdate.create(Scope scope, Operand<T> tensor, Operand<? extends TNumber> indices, Operand<T> updates, TensorScatterNdUpdate.Options... options) Factory method to create a class wrapping a new TensorScatterUpdate operation.static <T extends TType, U extends TNumber>
TensorStridedSliceUpdate<T> TensorStridedSliceUpdate.create(Scope scope, Operand<T> input, Operand<U> begin, Operand<U> end, Operand<U> strides, Operand<T> value, TensorStridedSliceUpdate.Options... options) Factory method to create a class wrapping a new TensorStridedSliceUpdate operation.Factory method to create a class wrapping a new Tile operation.static TopKUniqueFactory method to create a class wrapping a new TopKUnique operation.static TopKWithUniqueFactory method to create a class wrapping a new TopKWithUnique operation.Unbatch.create(Scope scope, Operand<T> batchedTensor, Operand<TInt64> batchIndex, Operand<TInt64> id, Long timeoutMicros, Unbatch.Options... options) Factory method to create a class wrapping a new Unbatch operation.static <T extends TType>
UnbatchGrad<T> UnbatchGrad.create(Scope scope, Operand<T> originalInput, Operand<TInt64> batchIndex, Operand<T> grad, Operand<TInt64> id, UnbatchGrad.Options... options) Factory method to create a class wrapping a new UnbatchGrad operation.static <T extends TNumber>
UniformQuantizedClipByValue<T> UniformQuantizedClipByValue.create(Scope scope, Operand<T> operand, Operand<T> min, Operand<T> max, Operand<TFloat32> scales, Operand<TInt32> zeroPoints, Long quantizationMinVal, Long quantizationMaxVal, UniformQuantizedClipByValue.Options... options) Factory method to create a class wrapping a new UniformQuantizedClipByValue operation.Factory method to create a class wrapping a new UniqueV2 operation, with the default output types.Factory method to create a class wrapping a new UniqueV2 operation.static <T extends TType>
UniqueWithCounts<T, TInt32> Factory method to create a class wrapping a new UniqueWithCountsV2 operation, with the default output types.static <T extends TType, V extends TNumber>
UniqueWithCounts<T, V> UniqueWithCounts.create(Scope scope, Operand<T> x, Operand<? extends TNumber> axis, Class<V> outIdx) Factory method to create a class wrapping a new UniqueWithCountsV2 operation.static <T extends TNumber>
UnravelIndex<T> Factory method to create a class wrapping a new UnravelIndex operation.Unstack.create(Scope scope, Operand<T> value, Long num, Unstack.Options... options) Factory method to create a class wrapping a new Unpack operation.static <T extends TType>
UpperBound<TInt32> Factory method to create a class wrapping a new UpperBound operation, with the default output types.static <U extends TNumber, T extends TType>
UpperBound<U> Factory method to create a class wrapping a new UpperBound operation.static VariableShape<TInt32> Factory method to create a class wrapping a new VariableShape operation, with the default output types.static <T extends TNumber>
VariableShape<T> Factory method to create a class wrapping a new VariableShape operation.static VarIsInitializedOpFactory method to create a class wrapping a new VarIsInitializedOp operation.static WhereFactory method to create a class wrapping a new Where operation.Creates a zeroed tensor given its type and shape.Factory method to create a class wrapping a new ZerosLike operation.Helpers.createVariableWithInit(Scope scope, Operand<T> init, Variable.Options... options) Factory method to create a new Variable with its initializer.Flatten the operand to 1 dimension.Flatten the operand to 1 dimensionCreates a 1-dimensional operand that represents a new shape containing the dimensions of an operand representing the shape to prepend, followed by the dimensions of an operand representing a shape.Reduces the shape to the specified axis.Shapes.reduceDims(Scope scope, Shape<U> shape, Operand<U> axis, Class<U> type) Reduces the shape to the specified axis.Shapes.reduceDims(Scope scope, Operand<T> operand, Operand<TInt32> axis) Reshapes the operand by reducing the shape to the specified axis.Shapes.reduceDims(Scope scope, Operand<T> operand, Operand<U> axis, Class<U> type) Reshapes the operand by reducing the shape to the specified axis.Get the size of the specified dimension in the shape.Get the size of the specified dimension in the shape.Get the size of the specified dimension for the shape of the tensor.Get the size of the specified dimension for the shape of the tensor.static <T extends TType>
StridedSlice<T> StridedSliceHelper.stridedSlice(Scope scope, Operand<T> input, Index... indices) Return a strided slice from `input`.static <T extends TType>
StridedSliceAssign<T> StridedSliceHelper.stridedSliceAssign(Scope scope, Operand<T> ref, Operand<T> value, Index... indices) Assign `value` to the sliced l-value reference of `ref`.Creates a 1-dimensional operand with the dimensions matching the first n dimensions of the shape.Creates a 1-dimensional operand containin the dimensions matching the first n dimensions of the shape.Creates a 1-dimensional operand containing the dimensions matching the last n dimensions of the shape.Creates a 1-dimensional operand containing the dimensions matching the last n dimensions of the shape.Constant.tensorOfSameType(Scope scope, Operand<T> toMatch, Number number) Creates a scalar of the same type astoMatch, with the value ofnumber.Method parameters in org.tensorflow.op.core with type arguments of type OperandModifier and TypeMethodDescriptionCalls the function in an execution environment, adding its graph as a function if it isn't already present.static AssertThatAssertThat.create(Scope scope, Operand<TBool> condition, Iterable<Operand<?>> data, AssertThat.Options... options) Factory method to create a class wrapping a new Assert operation.static BatchBatch.create(Scope scope, Iterable<Operand<?>> inTensors, Long numBatchThreads, Long maxBatchSize, Long batchTimeoutMicros, Long gradTimeoutMicros, Batch.Options... options) Factory method to create a class wrapping a new Batch operation.static BatchFunctionBatchFunction.create(Scope scope, Iterable<Operand<?>> inTensors, Iterable<Operand<?>> capturedTensors, ConcreteFunction f, Long numBatchThreads, Long maxBatchSize, Long batchTimeoutMicros, List<Class<? extends TType>> Tout, BatchFunction.Options... options) Factory method to create a class wrapping a new BatchFunction operation.static CaseCase.create(Scope scope, Operand<TInt32> branchIndex, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, List<ConcreteFunction> branches, Case.Options... options) Factory method to create a class wrapping a new Case operation.CompositeTensorVariantFromComponents.create(Scope scope, Iterable<Operand<?>> components, String metadata) Factory method to create a class wrapping a new CompositeTensorVariantFromComponents operation.Factory method to create a class wrapping a new ConcatV2 operation.static <T extends TNumber>
ConcatOffset<T> Factory method to create a class wrapping a new ConcatOffset operation.static <T extends TType>
DynamicStitch<T> Factory method to create a class wrapping a new DynamicStitch operation.static EncodeProtoEncodeProto.create(Scope scope, Operand<TInt32> sizes, Iterable<Operand<?>> values, List<String> fieldNames, String messageType, EncodeProto.Options... options) Factory method to create a class wrapping a new EncodeProto operation.static ForFor.create(Scope scope, Operand<TInt32> start, Operand<TInt32> limit, Operand<TInt32> delta, Iterable<Operand<?>> input, ConcreteFunction body) Factory method to create a class wrapping a new For operation.static GradientsGradients.create(Scope scope, Iterable<? extends Operand<?>> y, Iterable<? extends Operand<?>> x, Gradients.Options... options) Adds gradients computation ops to the graph according to scope.static GradientsGradients.create(Scope scope, Operand<?> y, Iterable<? extends Operand<?>> x, Gradients.Options... options) Adds gradients computation ops to the graph according to scope.static IdentityNFactory method to create a class wrapping a new IdentityN operation.static IfIf.create(Scope scope, Operand<? extends TType> cond, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, ConcreteFunction thenBranch, ConcreteFunction elseBranch, If.Options... options) Factory method to create a class wrapping a new If operation.static MapDefunMapDefun.create(Scope scope, Iterable<Operand<?>> arguments, Iterable<Operand<?>> capturedInputs, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ConcreteFunction f, MapDefun.Options... options) Factory method to create a class wrapping a new MapDefun operation.static MapStageMapStage.create(Scope scope, Operand<TInt64> key, Operand<TInt32> indices, Iterable<Operand<?>> values, List<Class<? extends TType>> dtypes, MapStage.Options... options) Factory method to create a class wrapping a new MapStage operation.Factory method to create a class wrapping a new Merge operation.static MlirPassthroughOpMlirPassthroughOp.create(Scope scope, Iterable<Operand<?>> inputs, String mlirModule, List<Class<? extends TType>> Toutputs) Factory method to create a class wrapping a new MlirPassthroughOp operation.static <T extends TNumber>
NcclReduce<T> Deprecated.Factory method to create a class wrapping a new NcclReduce operation.static OrderedMapStageOrderedMapStage.create(Scope scope, Operand<TInt64> key, Operand<TInt32> indices, Iterable<Operand<?>> values, List<Class<? extends TType>> dtypes, OrderedMapStage.Options... options) Factory method to create a class wrapping a new OrderedMapStage operation.static <T extends TType>
ParallelConcat<T> Factory method to create a class wrapping a new ParallelConcat operation.static <T extends TType>
ParallelDynamicStitch<T> ParallelDynamicStitch.create(Scope scope, Iterable<Operand<TInt32>> indices, Iterable<Operand<T>> data) Factory method to create a class wrapping a new ParallelDynamicStitch operation.static PartitionedCallPartitionedCall.create(Scope scope, Iterable<Operand<?>> args, List<Class<? extends TType>> Tout, ConcreteFunction f, PartitionedCall.Options... options) Factory method to create a class wrapping a new PartitionedCall operation.Factory method to create a class wrapping a new RefMerge operation.Factory method to create a class wrapping a new RefSelect operation.static RemoteCallRemoteCall.create(Scope scope, Operand<TString> target, Iterable<Operand<?>> args, List<Class<? extends TType>> Tout, ConcreteFunction f) Factory method to create a class wrapping a new RemoteCall operation.Factory method to create a class wrapping a new ShapeN operation, with the default output types.Factory method to create a class wrapping a new ShapeN operation.Stack.create(Scope scope, Iterable<Operand<T>> values, Stack.Options... options) Factory method to create a class wrapping a new Pack operation.static StageStage.create(Scope scope, Iterable<Operand<?>> values, Stage.Options... options) Factory method to create a class wrapping a new Stage operation.static StatefulCaseStatefulCase.create(Scope scope, Operand<TInt32> branchIndex, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, List<ConcreteFunction> branches, Case.Options... options) Factory method to create a class wrapping a new Case operation.static StatefulIfStatefulIf.create(Scope scope, Operand<? extends TType> cond, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, ConcreteFunction thenBranch, ConcreteFunction elseBranch, If.Options... options) Factory method to create a class wrapping a new If operation.static StatefulPartitionedCallStatefulPartitionedCall.create(Scope scope, Iterable<Operand<?>> args, List<Class<? extends TType>> Tout, ConcreteFunction f, StatefulPartitionedCall.Options... options) Factory method to create a class wrapping a new StatefulPartitionedCall operation.static StatefulWhileStatefulWhile.create(Scope scope, Iterable<Operand<?>> input, ConcreteFunction cond, ConcreteFunction body, While.Options... options) Factory method to create a class wrapping a new While operation.static StatelessCaseStatelessCase.create(Scope scope, Operand<TInt32> branchIndex, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, List<ConcreteFunction> branches, Case.Options... options) Factory method to create a class wrapping a new StatelessCase operation.static StatelessIfStatelessIf.create(Scope scope, Operand<? extends TType> cond, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, ConcreteFunction thenBranch, ConcreteFunction elseBranch, If.Options... options) Factory method to create a class wrapping a new StatelessIf operation.static StatelessWhileStatelessWhile.create(Scope scope, Iterable<Operand<?>> input, ConcreteFunction cond, ConcreteFunction body, While.Options... options) Factory method to create a class wrapping a new StatelessWhile operation.static WhileWhile.create(Scope scope, Iterable<Operand<?>> input, ConcreteFunction cond, ConcreteFunction body, While.Options... options) Factory method to create a class wrapping a new While operation.static Gradients.Options -
Uses of Operand in org.tensorflow.op.data
Classes in org.tensorflow.op.data that implement OperandModifier and TypeClassDescriptionfinal classA container for an iterator resource.final classA container for a multi device iterator resource.final classThe AssertCardinalityDataset operationfinal classA transformation that asserts which transformations happen next.final classA transformation that asserts which transformations happened previously.final classCreates a dataset that shards the input dataset.final classCreates a dataset that batchesbatch_sizeelements frominput_dataset.final classRecords the bytes size of each element ofinput_datasetin a StatsAggregator.final classThe CacheDatasetV2 operationfinal classThe ChooseFastestBranchDataset operationfinal classThe ChooseFastestDataset operationfinal classCompresses a dataset element.final classCreates a dataset that concatenatesinput_datasetwithanother_dataset.final classThe CSVDatasetV2 operationfinal classCreates a dataset that reads data from the tf.data service.final classReturns the cardinality ofinput_dataset.final classReturns the fingerprint ofinput_dataset.final classCreates a dataset from the givengraph_def.final classReturns a serialized GraphDef representinginput_dataset.final classCreates a dataset that batches input elements into a SparseTensor.final classA substitute forInterleaveDataseton a fixed list ofNdatasets.final classThe DummyIterationCounter operationfinal classCreates a dataset containing elements of first component ofinput_datasethaving true in the last component.final classCreates a dataset containing elements ofinput_datasetmatchingpredicate.final classCreates a dataset by applyingtf.data.Optionstoinput_dataset.final classThe FixedLengthRecordDatasetV2 operationfinal classCreates a dataset that appliesfto the outputs ofinput_dataset.final classCreates a dataset that invokes a function to generate elements.final classThe GlobalShuffleDataset operationfinal classCreates a dataset that computes a group-by oninput_dataset.final classCreates a dataset that computes a windowed group-by oninput_dataset. // TODO(mrry): Support non-int64 keys.final classCreates a dataset that contains the elements ofinput_datasetignoring errors.final classThe IndexFlatMapDataset operationfinal classCreates a dataset that appliesfto the outputs ofinput_dataset.final classThe IteratorV2 operationfinal classThe IteratorFromStringHandleV2 operationfinal classReturns the name of the device on whichresourcehas been placed.final classReturns the serialized model proto of an iterator resource.final classGets the next output from the given iterator as an Optional variant.final classConverts the givenresource_handlerepresenting an iterator to a string.final classRecords the latency of producinginput_datasetelements in a StatsAggregator.final classLeakyReluGrad<T extends TNumber>Computes rectified linear gradients for a LeakyRelu operation.final classCreates a dataset that appliesfto the outputs ofinput_dataset.final classCreates a dataset that emits each oftensorsonce.final classThe ListSnapshotChunksDataset operationfinal classCreates a dataset that emits the key-value pairs in one or more LMDB files.final classThe LoadDataset operationfinal classCreates a dataset that fuses mapping with batching.final classCreates a dataset that appliesfto the outputs ofinput_dataset.final classThe MatchingFilesDataset operationfinal classCreates a dataset that overrides the maximum intra-op parallelism.final classIdentity transformation that models performance.final classCreates a MultiDeviceIterator resource.final classGenerates a MultiDeviceIterator resource from its provided string handle.final classInitializes the multi device iterator with the given dataset.final classProduces a string handle for the given MultiDeviceIterator.final classThe NonSerializableDataset operationfinal classMakes a "one-shot" iterator that can be iterated only once.final classCreates a dataset by applying related optimizations toinput_dataset.final classConstructs an Optional variant from a tuple of tensors.final classReturns true if and only if the given Optional variant has a value.final classCreates an Optional variant with no value.final classCreates a dataset by attaching tf.data.Options toinput_dataset.final classCreates a dataset that batches and padsbatch_sizeelements from the input.final classThe ParallelBatchDataset operationfinal classCreates a dataset containing elements ofinput_datasetmatchingpredicate.final classCreates a dataset that appliesfto the outputs ofinput_dataset.final classCreates a dataset that appliesfto the outputs ofinput_dataset.final classTransformsinput_datasetcontainingExampleprotos as vectors of DT_STRING into a dataset ofTensororSparseTensorobjects representing the parsed features.final classCreates a dataset that asynchronously prefetches elements frominput_dataset.final classCreates a dataset that uses a custom thread pool to computeinput_dataset.final classCreates a Dataset that returns pseudorandom numbers.final classCreates a dataset with a range of values.final classCreates a dataset that changes the batch size.final classRegisters a dataset with the tf.data service.final classCreates a dataset that emits the outputs ofinput_datasetcounttimes.final classThe RewriteDataset operationfinal classCreates a dataset that takes a Bernoulli sample of the contents of another dataset.final classThe SaveDatasetV2 operationfinal classCreates a dataset successively reducesfover the elements ofinput_dataset.final classConverts the givenresource_handlerepresenting an iterator to a variant tensor.final classThe SetStatsAggregatorDataset operationfinal classCreates aDatasetthat includes only 1/num_shardsof this dataset.final classThe ShuffleAndRepeatDatasetV2 operationfinal classThe ShuffleDatasetV3 operationfinal classCreates a dataset that skipscountelements from theinput_dataset.final classThe SleepDataset operationfinal classCreates a dataset that passes a sliding window overinput_dataset.final classThe SnapshotChunkDataset operationfinal classCreates a dataset that will write to / read from a snapshot.final classThe SnapshotDatasetReader operationfinal classThe SnapshotNestedDatasetReader operationfinal classCreates a dataset that splits a SparseTensor into elements row-wise.final classCreates a dataset that executes a SQL query and emits rows of the result set.final classThe StatsAggregatorHandleV2 operationfinal classCreates a dataset that containscountelements from theinput_dataset.final classCreates a dataset that stops iteration when predicate` is false.final classCreates a dataset that emitscomponentsas a tuple of tensors once.final classCreates a dataset that emits each dim-0 slice ofcomponentsonce.final classCreates a dataset that emits the lines of one or more text files.final classCreates a dataset that emits the records from one or more TFRecord files.final classCreates a dataset that uses a custom thread pool to computeinput_dataset.final classCreates a dataset that uses a custom thread pool to computeinput_dataset.final classA dataset that splits the elements of its input into multiple elements.final classCreates a dataset that contains the unique elements ofinput_dataset.final classThe UnwrapDatasetVariant operationfinal classThe WeightedFlatMapDataset operationfinal classCombines (nests of) input elements into a dataset of (nests of) windows.final classThe WindowOp operationfinal classThe WrapDatasetVariant operationfinal classCreates a dataset that zips togetherinput_datasets.Classes in org.tensorflow.op.data that implement interfaces with type arguments of type OperandModifier and TypeClassDescriptionfinal classOutputs the single element from the given dataset.final classGets the next output from the given iterator .final classGets the next output from the given iterator.final classGets next element for the provided shard number.final classReturns the value stored in an Optional variant or raises an error if none exists.final classReduces the input dataset to a singleton using a reduce function.final classUncompresses a compressed dataset element.Fields in org.tensorflow.op.data declared as OperandModifier and TypeFieldDescriptionDataServiceDataset.Inputs.addressThe address inputRegisterDataset.Inputs.addressThe address inputConcatenateDataset.Inputs.anotherDatasetThe anotherDataset inputBatchDataset.Inputs.batchSizeA scalar representing the number of elements to accumulate in a batch.DenseToSparseBatchDataset.Inputs.batchSizeA scalar representing the number of elements to accumulate in a batch.MapAndBatchDataset.Inputs.batchSizeA scalar representing the number of elements to accumulate in a batch.PaddedBatchDataset.Inputs.batchSizeA scalar representing the number of elements to accumulate in a batch.ParallelBatchDataset.Inputs.batchSizeThe batchSize inputRebatchDatasetV2.Inputs.batchSizesA vector of integers representing the size of batches to produce.InterleaveDataset.Inputs.blockLengthThe blockLength inputLegacyParallelInterleaveDataset.Inputs.blockLengthThe blockLength inputParallelInterleaveDataset.Inputs.blockLengthNumber of elements at a time to produce from each interleaved invocation of a dataset returned byf.LegacyParallelInterleaveDataset.Inputs.bufferOutputElementsThe bufferOutputElements inputParallelInterleaveDataset.Inputs.bufferOutputElementsThe number of elements each iterator being interleaved should buffer (similar to the.prefetch()transformation for each interleaved iterator).CSVDataset.Inputs.bufferSizeThe bufferSize inputFixedLengthRecordDataset.Inputs.bufferSizeThe bufferSize inputPrefetchDataset.Inputs.bufferSizeThe maximum number of elements to buffer in an iterator over this dataset.ShuffleAndRepeatDataset.Inputs.bufferSizeThe bufferSize inputShuffleDataset.Inputs.bufferSizeThe bufferSize inputTextLineDataset.Inputs.bufferSizeA scalar containing the number of bytes to buffer.TfRecordDataset.Inputs.bufferSizeA scalar representing the number of bytes to buffer.TfRecordDataset.Inputs.byteOffsetsA scalar or vector containing the number of bytes for each file that will be skipped prior to reading.CacheDataset.Inputs.cacheThe cache inputAssertCardinalityDataset.Inputs.cardinalityThe cardinality inputSnapshotChunkDataset.Inputs.chunkFileThe chunkFile inputUncompressElement.Inputs.compressedThe compressed inputCSVDataset.Inputs.compressionTypeThe compressionType inputDatasetToTfRecord.Inputs.compressionTypeA scalar string tensor containing either (i) the empty string (no compression), (ii) "ZLIB", or (iii) "GZIP".FixedLengthRecordDataset.Inputs.compressionTypeThe compressionType inputTextLineDataset.Inputs.compressionTypeA scalar containing either (i) the empty string (no compression), (ii) "ZLIB", or (iii) "GZIP".TfRecordDataset.Inputs.compressionTypeA scalar containing either (i) the empty string (no compression), (ii) "ZLIB", or (iii) "GZIP".DataServiceDataset.Inputs.consumerIndexThe consumerIndex inputRepeatDataset.Inputs.countA scalar representing the number of times thatinput_datasetshould be repeated.ShuffleAndRepeatDataset.Inputs.countThe count inputSkipDataset.Inputs.countA scalar representing the number of elements from theinput_datasetthat should be skipped.TakeDataset.Inputs.countA scalar representing the number of elements from theinput_datasetthat should be taken.SetStatsAggregatorDataset.Inputs.counterPrefixThe counterPrefix inputInterleaveDataset.Inputs.cycleLengthThe cycleLength inputLegacyParallelInterleaveDataset.Inputs.cycleLengthThe cycleLength inputParallelInterleaveDataset.Inputs.cycleLengthNumber of datasets (each created by applyingfto the elements ofinput_dataset) among which theParallelInterleaveDatasetV2will cycle in a round-robin fashion.DatasetToSingleElement.Inputs.datasetA handle to a dataset that contains a single element.InitializeTableFromDataset.Inputs.datasetThe dataset inputMakeIterator.Inputs.datasetThe dataset inputMultiDeviceIteratorInit.Inputs.datasetDataset to be iterated upon.RegisterDataset.Inputs.datasetThe dataset inputDataServiceDataset.Inputs.datasetIdThe datasetId inputSqlDataset.Inputs.dataSourceNameA connection string to connect to the database.DeleteIterator.Inputs.deleterA variant deleter.DeleteMemoryCache.Inputs.deleterThe deleter inputDeleteMultiDeviceIterator.Inputs.deleterA variant deleter.SparseTensorSliceDataset.Inputs.denseShapeThe denseShape inputSqlDataset.Inputs.driverNameThe database type.BatchDataset.Inputs.dropRemainderA scalar representing whether the last batch should be dropped in case its size is smaller than desired.MapAndBatchDataset.Inputs.dropRemainderA scalar representing whether the last batch should be dropped in case its size is smaller than desired.PaddedBatchDataset.Inputs.dropRemainderA scalar representing whether the last batch should be dropped in case its size is smaller than desired.ParallelBatchDataset.Inputs.dropRemainderThe dropRemainder inputRebatchDatasetV2.Inputs.dropRemainderThe dropRemainder inputWindowDataset.Inputs.dropRemainderA Boolean scalar, representing whether the last window should be dropped if its size is smaller thanwindow_size.CSVDataset.Inputs.excludeColsThe excludeCols inputLeakyReluGrad.Inputs.featuresThe features passed as input to the corresponding LeakyRelu operation, OR the outputs of that operation (both work equivalently).CSVDataset.Inputs.fieldDelimThe fieldDelim inputCacheDataset.Inputs.filenameThe filename inputDatasetToTfRecord.Inputs.filenameA scalar string tensor representing the filename to use.CSVDataset.Inputs.filenamesThe filenames inputFixedLengthRecordDataset.Inputs.filenamesThe filenames inputLMDBDataset.Inputs.filenamesA scalar or a vector containing the name(s) of the binary file(s) to be read.TextLineDataset.Inputs.filenamesA scalar or a vector containing the name(s) of the file(s) to be read.TfRecordDataset.Inputs.filenamesA scalar or vector containing the name(s) of the file(s) to be read.FixedLengthRecordDataset.Inputs.footerBytesThe footerBytes inputLeakyReluGrad.Inputs.gradientsThe backpropagated gradients to the corresponding LeakyRelu operation.DatasetFromGraph.Inputs.graphDefThe graph representation of the dataset (as serialized GraphDef).DeleteIterator.Inputs.handleA handle to the iterator to delete.DeleteMemoryCache.Inputs.handleThe handle inputCSVDataset.Inputs.headerThe header inputFixedLengthRecordDataset.Inputs.headerBytesThe headerBytes inputMultiDeviceIteratorGetNextFromShard.Inputs.incarnationIdWhich incarnation of the MultiDeviceIterator is running.AutoShardDataset.Inputs.indexA scalar representing the index of the current worker out of num_workers.ShardDataset.Inputs.indexAn integer representing the current worker index.SparseTensorSliceDataset.Inputs.indicesThe indices inputAssertCardinalityDataset.Inputs.inputDatasetThe inputDataset inputAssertNextDataset.Inputs.inputDatasetA variant tensor representing the input dataset.AssertPrevDataset.Inputs.inputDatasetA variant tensor representing the input dataset.AutoShardDataset.Inputs.inputDatasetA variant tensor representing the input dataset.BatchDataset.Inputs.inputDatasetThe inputDataset inputBytesProducedStatsDataset.Inputs.inputDatasetThe inputDataset inputCacheDataset.Inputs.inputDatasetThe inputDataset inputChooseFastestBranchDataset.Inputs.inputDatasetThe inputDataset inputConcatenateDataset.Inputs.inputDatasetThe inputDataset inputDatasetCardinality.Inputs.inputDatasetA variant tensor representing the dataset to return cardinality for.DatasetFingerprint.Inputs.inputDatasetA variant tensor representing the dataset to return fingerprint for.DatasetToGraph.Inputs.inputDatasetA variant tensor representing the dataset to return the graph representation for.DatasetToTfRecord.Inputs.inputDatasetA variant tensor representing the dataset to write.DenseToSparseBatchDataset.Inputs.inputDatasetA handle to an input dataset.FilterByLastComponentDataset.Inputs.inputDatasetThe inputDataset inputFilterDataset.Inputs.inputDatasetThe inputDataset inputFinalizeDataset.Inputs.inputDatasetA variant tensor representing the input dataset.FlatMapDataset.Inputs.inputDatasetThe inputDataset inputGlobalShuffleDataset.Inputs.inputDatasetThe inputDataset inputGroupByReducerDataset.Inputs.inputDatasetA variant tensor representing the input dataset.GroupByWindowDataset.Inputs.inputDatasetThe inputDataset inputIgnoreErrorsDataset.Inputs.inputDatasetThe inputDataset inputIndexFlatMapDataset.Inputs.inputDatasetThe inputDataset inputInterleaveDataset.Inputs.inputDatasetThe inputDataset inputLatencyStatsDataset.Inputs.inputDatasetThe inputDataset inputLegacyParallelInterleaveDataset.Inputs.inputDatasetThe inputDataset inputMapAndBatchDataset.Inputs.inputDatasetA variant tensor representing the input dataset.MapDataset.Inputs.inputDatasetThe inputDataset inputMaxIntraOpParallelismDataset.Inputs.inputDatasetThe inputDataset inputModelDataset.Inputs.inputDatasetA variant tensor representing the input dataset.NonSerializableDataset.Inputs.inputDatasetThe inputDataset inputOptimizeDataset.Inputs.inputDatasetA variant tensor representing the input dataset.OptionsDataset.Inputs.inputDatasetA variant tensor representing the input dataset.PaddedBatchDataset.Inputs.inputDatasetThe inputDataset inputParallelBatchDataset.Inputs.inputDatasetThe inputDataset inputParallelFilterDataset.Inputs.inputDatasetThe inputDataset inputParallelInterleaveDataset.Inputs.inputDatasetDataset that produces a stream of arguments for the functionf.ParallelMapDataset.Inputs.inputDatasetThe inputDataset inputParseExampleDataset.Inputs.inputDatasetThe inputDataset inputPrefetchDataset.Inputs.inputDatasetThe inputDataset inputPrivateThreadPoolDataset.Inputs.inputDatasetThe inputDataset inputRebatchDatasetV2.Inputs.inputDatasetA variant tensor representing the input dataset.ReduceDataset.Inputs.inputDatasetA variant tensor representing the input dataset.RepeatDataset.Inputs.inputDatasetThe inputDataset inputRewriteDataset.Inputs.inputDatasetThe inputDataset inputSamplingDataset.Inputs.inputDatasetThe inputDataset inputSaveDataset.Inputs.inputDatasetThe inputDataset inputScanDataset.Inputs.inputDatasetThe inputDataset inputSetStatsAggregatorDataset.Inputs.inputDatasetThe inputDataset inputShardDataset.Inputs.inputDatasetThe inputDataset inputShuffleAndRepeatDataset.Inputs.inputDatasetThe inputDataset inputShuffleDataset.Inputs.inputDatasetThe inputDataset inputSkipDataset.Inputs.inputDatasetThe inputDataset inputSleepDataset.Inputs.inputDatasetThe inputDataset inputSlidingWindowDataset.Inputs.inputDatasetThe inputDataset inputSnapshotDataset.Inputs.inputDatasetA variant tensor representing the input dataset.TakeDataset.Inputs.inputDatasetThe inputDataset inputTakeWhileDataset.Inputs.inputDatasetThe inputDataset inputThreadPoolDataset.Inputs.inputDatasetThe inputDataset inputUnbatchDataset.Inputs.inputDatasetThe inputDataset inputUniqueDataset.Inputs.inputDatasetThe inputDataset inputWindowDataset.Inputs.inputDatasetThe inputDataset inputUnwrapDatasetVariant.Inputs.inputHandleThe inputHandle inputWrapDatasetVariant.Inputs.inputHandleThe inputHandle inputDataServiceDataset.Inputs.iterationCounterThe iterationCounter inputIteratorGetModelProto.Inputs.iteratorAn resource from an dataset iterator.IteratorGetNext.Inputs.iteratorThe iterator inputIteratorGetNextAsOptional.Inputs.iteratorThe iterator inputIteratorGetNextSync.Inputs.iteratorThe iterator inputMakeIterator.Inputs.iteratorThe iterator inputDataServiceDataset.Inputs.jobNameThe jobName inputMultiDeviceIteratorInit.Inputs.maxBufferSizeThe maximum size of the host side per device buffer to keep.MaxIntraOpParallelismDataset.Inputs.maxIntraOpParallelismIdentifies the maximum intra-op parallelism to use.DataServiceDataset.Inputs.maxOutstandingRequestsThe maxOutstandingRequests inputDeleteMultiDeviceIterator.Inputs.multiDeviceIteratorA handle to the multi device iterator to delete.MultiDeviceIteratorGetNextFromShard.Inputs.multiDeviceIteratorA MultiDeviceIterator resource.MultiDeviceIteratorInit.Inputs.multiDeviceIteratorA MultiDeviceIteratorResource.MultiDeviceIteratorToStringHandle.Inputs.multiDeviceIteratorA MultiDeviceIterator resource.CSVDataset.Inputs.naValueThe naValue inputDataServiceDataset.Inputs.numConsumersThe numConsumers inputMapAndBatchDataset.Inputs.numParallelCallsA scalar representing the maximum number of parallel invocations of themap_fnfunction.ParallelBatchDataset.Inputs.numParallelCallsThe numParallelCalls inputParallelFilterDataset.Inputs.numParallelCallsThe number of concurrent invocations ofpredicatethat process elements frominput_datasetin parallel.ParallelInterleaveDataset.Inputs.numParallelCallsDetermines the number of threads that should be used for fetching data from input datasets in parallel.ParallelMapDataset.Inputs.numParallelCallsThe number of concurrent invocations offthat process elements frominput_datasetin parallel.ParseExampleDataset.Inputs.numParallelCallsThe numParallelCalls inputShardDataset.Inputs.numShardsAn integer representing the number of shards operating in parallel.PrivateThreadPoolDataset.Inputs.numThreadsIdentifies the number of threads to use for the private threadpool.AutoShardDataset.Inputs.numWorkersA scalar representing the number of workers to distribute this dataset across.OptimizeDataset.Inputs.optimizationsDefaultAtf.stringvectortf.Tensoridentifying optimizations by default.OptimizeDataset.Inputs.optimizationsDisabledAtf.stringvectortf.Tensoridentifying user disabled optimizations.OptimizeDataset.Inputs.optimizationsEnabledAtf.stringvectortf.Tensoridentifying user enabled optimizations.OptionalGetValue.Inputs.optionalThe optional inputOptionalHasValue.Inputs.optionalThe optional inputIndexFlatMapDataset.Inputs.outputCardinalityThe outputCardinality inputLoadDataset.Inputs.pathThe path inputSaveDataset.Inputs.pathThe path inputSnapshotDataset.Inputs.pathThe path we should write snapshots to / read snapshots from.MatchingFilesDataset.Inputs.patternsThe patterns inputLegacyParallelInterleaveDataset.Inputs.prefetchInputElementsThe prefetchInputElements inputParallelInterleaveDataset.Inputs.prefetchInputElementsDetermines the number of iterators to prefetch, allowing buffers to warm up and data to be pre-fetched without blocking the main thread.DataServiceDataset.Inputs.processingModeThe processingMode inputDataServiceDataset.Inputs.protocolThe protocol inputRegisterDataset.Inputs.protocolThe protocol inputSqlDataset.Inputs.queryA SQL query to execute.SamplingDataset.Inputs.rateA scalar representing the sample rate.ChooseFastestBranchDataset.Inputs.ratioDenominatorThe ratioDenominator inputChooseFastestBranchDataset.Inputs.ratioNumeratorThe ratioNumerator inputFixedLengthRecordDataset.Inputs.recordBytesThe recordBytes inputIteratorGetDevice.Inputs.resourceThe resource inputDeserializeIterator.Inputs.resourceHandleA handle to an iterator resource.IteratorToStringHandle.Inputs.resourceHandleA handle to an iterator resource.SerializeIterator.Inputs.resourceHandleA handle to an iterator resource.RewriteDataset.Inputs.rewriteNameThe rewriteName inputDenseToSparseBatchDataset.Inputs.rowShapeA vector representing the dense shape of each row in the produced SparseTensor.GlobalShuffleDataset.Inputs.seedThe seed inputRandomDataset.Inputs.seedA scalar seed for the random number generator.SamplingDataset.Inputs.seedA scalar representing seed of random number generator.ShuffleAndRepeatDataset.Inputs.seedThe seed inputShuffleDataset.Inputs.seedThe seed inputGlobalShuffleDataset.Inputs.seed2The seed2 inputRandomDataset.Inputs.seed2A second scalar seed to avoid seed collision.SamplingDataset.Inputs.seed2A scalar representing seed2 of random number generator.ShuffleAndRepeatDataset.Inputs.seed2The seed2 inputShuffleDataset.Inputs.seed2The seed2 inputGlobalShuffleDataset.Inputs.seedGeneratorThe seedGenerator inputRandomDataset.Inputs.seedGeneratorA resource for the random number seed generator.ShuffleAndRepeatDataset.Inputs.seedGeneratorThe seedGenerator inputShuffleDataset.Inputs.seedGeneratorThe seedGenerator inputCSVDataset.Inputs.selectColsThe selectCols inputDirectedInterleaveDataset.Inputs.selectorInputDatasetA dataset of scalarDT_INT64elements that determines which of theNdata inputs should produce the next output element.DeserializeIterator.Inputs.serializedA variant tensor storing the state of the iterator contained in the resource.SnapshotDatasetReader.Inputs.shardDirThe shardDir inputMultiDeviceIteratorGetNextFromShard.Inputs.shardNumInteger representing which shard to fetch data for.WindowDataset.Inputs.shiftAn integer scalar, representing the number of input elements by which the window moves in each iteration.WindowDataset.Inputs.sizeOutputAn integer scalar, representing the number of elements of the input dataset to combine into a window.SleepDataset.Inputs.sleepMicrosecondsThe sleepMicroseconds inputListSnapshotChunksDataset.Inputs.snapshotPathThe snapshotPath inputRangeDataset.Inputs.startcorresponds to start in python's xrange().SnapshotDatasetReader.Inputs.startIndexThe startIndex inputSetStatsAggregatorDataset.Inputs.statsAggregatorThe statsAggregator inputStatsAggregatorSetSummaryWriter.Inputs.statsAggregatorThe statsAggregator inputRangeDataset.Inputs.stepcorresponds to step in python's xrange().RangeDataset.Inputs.stopcorresponds to stop in python's xrange().WindowDataset.Inputs.strideAn integer scalar, representing the stride of the input elements in the sliding window.IteratorFromStringHandle.Inputs.stringHandleThe stringHandle inputMultiDeviceIteratorFromStringHandle.Inputs.stringHandleString representing the resource.StatsAggregatorSetSummaryWriter.Inputs.summaryThe summary inputInitializeTableFromDataset.Inputs.tableHandleThe tableHandle inputBytesProducedStatsDataset.Inputs.tagThe tag inputLatencyStatsDataset.Inputs.tagThe tag inputSetStatsAggregatorDataset.Inputs.tagThe tag inputThreadPoolDataset.Inputs.threadPoolA resource produced by the ThreadPoolHandle op.AssertNextDataset.Inputs.transformationsAtf.stringvectortf.Tensoridentifying the transformations that are expected to happen next.AssertPrevDataset.Inputs.transformationsAtf.stringvectortf.Tensoridentifying the transformations, with optional attribute name-value pairs, that are expected to have happened previously.CSVDataset.Inputs.useQuoteDelimThe useQuoteDelim inputSparseTensorSliceDataset.Inputs.valuesThe values inputSlidingWindowDataset.Inputs.windowShiftA scalar representing the steps moving the sliding window forward in one iteration.SlidingWindowDataset.Inputs.windowSizeA scalar representing the number of elements in the sliding window.SlidingWindowDataset.Inputs.windowStrideA scalar representing the stride of the input elements of the sliding window.Fields in org.tensorflow.op.data with type parameters of type OperandModifier and TypeFieldDescriptionCompressElement.Inputs.componentsThe components inputOptionalFromValue.Inputs.componentsThe components inputTensorDataset.Inputs.componentsThe components inputTensorSliceDataset.Inputs.componentsThe components inputDirectedInterleaveDataset.Inputs.dataInputDatasetsNdatasets with the same type that will be interleaved according to the values ofselector_input_dataset.ParseExampleDataset.Inputs.denseDefaultsA dict mapping string keys toTensors.GeneratorDataset.Inputs.finalizeFuncOtherArgsThe finalizeFuncOtherArgs inputGroupByReducerDataset.Inputs.finalizeFuncOtherArgumentsA list of tensors, typically values that were captured when building a closure forfinalize_func.IndexFlatMapDataset.Inputs.indexMapFuncOtherArgsThe indexMapFuncOtherArgs inputGeneratorDataset.Inputs.initFuncOtherArgsThe initFuncOtherArgs inputGroupByReducerDataset.Inputs.initFuncOtherArgumentsA list of tensors, typically values that were captured when building a closure forinit_func.ReduceDataset.Inputs.initialStateA nested structure of tensors, representing the initial state of the transformation.ScanDataset.Inputs.initialStateThe initialState inputChooseFastestDataset.Inputs.inputDatasetsThe inputDatasets inputWeightedFlatMapDataset.Inputs.inputDatasetsThe inputDatasets inputZipDataset.Inputs.inputDatasetsList ofNvariant Tensors representing datasets to be zipped together.SnapshotNestedDatasetReader.Inputs.inputsThe inputs inputWindowOp.Inputs.inputsThe inputs inputDeleteMultiDeviceIterator.Inputs.iteratorsA list of iterator handles (unused).GroupByReducerDataset.Inputs.keyFuncOtherArgumentsA list of tensors, typically values that were captured when building a closure forkey_func.GroupByWindowDataset.Inputs.keyFuncOtherArgumentsThe keyFuncOtherArguments inputIndexFlatMapDataset.Inputs.mapFuncOtherArgsThe mapFuncOtherArgs inputGeneratorDataset.Inputs.nextFuncOtherArgsThe nextFuncOtherArgs inputChooseFastestBranchDataset.Inputs.otherArgumentsThe otherArguments inputFilterDataset.Inputs.otherArgumentsA list of tensors, typically values that were captured when building a closure forpredicate.FlatMapDataset.Inputs.otherArgumentsThe otherArguments inputInterleaveDataset.Inputs.otherArgumentsThe otherArguments inputLegacyParallelInterleaveDataset.Inputs.otherArgumentsThe otherArguments inputMapAndBatchDataset.Inputs.otherArgumentsA list of tensors, typically values that were captured when building a closure forf.MapDataset.Inputs.otherArgumentsThe otherArguments inputParallelFilterDataset.Inputs.otherArgumentsA list of tensors, typically values that were captured when building a closure forpredicate.ParallelInterleaveDataset.Inputs.otherArgumentsAdditional arguments to pass tofbeyond those produced byinput_dataset.ParallelMapDataset.Inputs.otherArgumentsThe otherArguments inputReduceDataset.Inputs.otherArgumentsThe otherArguments inputScanDataset.Inputs.otherArgumentsThe otherArguments inputTakeWhileDataset.Inputs.otherArgumentsA list of tensors, typically values that were captured when building a closure forpredicate.PaddedBatchDataset.Inputs.paddedShapesA list of int64 tensors representing the desired padded shapes of the corresponding output components.PaddedBatchDataset.Inputs.paddingValuesA list of scalars containing the padding value to use for each of the outputs.LoadDataset.Inputs.readerFuncOtherArgsThe readerFuncOtherArgs inputSnapshotDataset.Inputs.readerFuncOtherArgsThe readerFuncOtherArgs inputCSVDataset.Inputs.recordDefaultsThe recordDefaults inputGroupByReducerDataset.Inputs.reduceFuncOtherArgumentsA list of tensors, typically values that were captured when building a closure forreduce_func.GroupByWindowDataset.Inputs.reduceFuncOtherArgumentsThe reduceFuncOtherArguments inputSaveDataset.Inputs.shardFuncOtherArgsThe shardFuncOtherArgs inputSnapshotDataset.Inputs.shardFuncOtherArgsThe shardFuncOtherArgs inputListDataset.Inputs.tensorsThe tensors inputWeightedFlatMapDataset.Inputs.weightsThe weights inputGroupByWindowDataset.Inputs.windowSizeFuncOtherArgumentsThe windowSizeFuncOtherArguments inputMethods in org.tensorflow.op.data that return types with arguments of type OperandModifier and TypeMethodDescriptionDatasetToSingleElement.iterator()IteratorGetNext.iterator()IteratorGetNextSync.iterator()MultiDeviceIteratorGetNextFromShard.iterator()OptionalGetValue.iterator()ReduceDataset.iterator()UncompressElement.iterator()Methods in org.tensorflow.op.data with parameters of type OperandModifier and TypeMethodDescriptionstatic AssertCardinalityDatasetAssertCardinalityDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> cardinality, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new AssertCardinalityDataset operation.static AssertNextDatasetAssertNextDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TString> transformations, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new AssertNextDataset operation.static AssertPrevDatasetAssertPrevDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TString> transformations, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new AssertPrevDataset operation.static AutoShardDatasetAutoShardDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> numWorkers, Operand<TInt64> index, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, AutoShardDataset.Options... options) Factory method to create a class wrapping a new AutoShardDataset operation.static BatchDatasetBatchDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> batchSize, Operand<TBool> dropRemainder, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, BatchDataset.Options... options) Factory method to create a class wrapping a new BatchDatasetV2 operation.static BytesProducedStatsDatasetBytesProducedStatsDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TString> tag, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new BytesProducedStatsDataset operation.static CacheDatasetCacheDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TString> filename, Operand<? extends TType> cache, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, CacheDataset.Options... options) Factory method to create a class wrapping a new CacheDatasetV2 operation.static ChooseFastestBranchDatasetChooseFastestBranchDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> ratioNumerator, Operand<TInt64> ratioDenominator, Iterable<Operand<?>> otherArguments, Long numElementsPerBranch, List<ConcreteFunction> branches, List<Long> otherArgumentsLengths, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ChooseFastestBranchDataset operation.static ConcatenateDatasetConcatenateDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<? extends TType> anotherDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ConcatenateDataset.Options... options) Factory method to create a class wrapping a new ConcatenateDataset operation.static CSVDatasetCSVDataset.create(Scope scope, Operand<TString> filenames, Operand<TString> compressionType, Operand<TInt64> bufferSize, Operand<TBool> header, Operand<TString> fieldDelim, Operand<TBool> useQuoteDelim, Operand<TString> naValue, Operand<TInt64> selectCols, Iterable<Operand<?>> recordDefaults, Operand<TInt64> excludeCols, List<Shape> outputShapes) Factory method to create a class wrapping a new CSVDatasetV2 operation.static DataServiceDatasetDataServiceDataset.create(Scope scope, Operand<TString> datasetId, Operand<TString> processingMode, Operand<TString> address, Operand<TString> protocol, Operand<TString> jobName, Operand<TInt64> consumerIndex, Operand<TInt64> numConsumers, Operand<TInt64> maxOutstandingRequests, Operand<? extends TType> iterationCounter, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ConcreteFunction uncompressFn, DataServiceDataset.Options... options) Factory method to create a class wrapping a new DataServiceDatasetV4 operation.static DatasetCardinalityDatasetCardinality.create(Scope scope, Operand<? extends TType> inputDataset, DatasetCardinality.Options... options) Factory method to create a class wrapping a new DatasetCardinality operation.static DatasetFingerprintFactory method to create a class wrapping a new DatasetFingerprint operation.static DatasetFromGraphFactory method to create a class wrapping a new DatasetFromGraph operation.static DatasetToGraphDatasetToGraph.create(Scope scope, Operand<? extends TType> inputDataset, DatasetToGraph.Options... options) Factory method to create a class wrapping a new DatasetToGraphV2 operation.static DatasetToSingleElementDatasetToSingleElement.create(Scope scope, Operand<? extends TType> dataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, DatasetToSingleElement.Options... options) Factory method to create a class wrapping a new DatasetToSingleElement operation.static DatasetToTfRecordDatasetToTfRecord.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TString> filename, Operand<TString> compressionType) Factory method to create a class wrapping a new DatasetToTFRecord operation.static DeleteIteratorDeleteIterator.create(Scope scope, Operand<? extends TType> handle, Operand<? extends TType> deleter) Factory method to create a class wrapping a new DeleteIterator operation.static DeleteMemoryCacheDeleteMemoryCache.create(Scope scope, Operand<? extends TType> handle, Operand<? extends TType> deleter) Factory method to create a class wrapping a new DeleteMemoryCache operation.static DeleteMultiDeviceIteratorDeleteMultiDeviceIterator.create(Scope scope, Operand<? extends TType> multiDeviceIterator, Iterable<Operand<? extends TType>> iterators, Operand<? extends TType> deleter) Factory method to create a class wrapping a new DeleteMultiDeviceIterator operation.static DenseToSparseBatchDatasetDenseToSparseBatchDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> batchSize, Operand<TInt64> rowShape, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new DenseToSparseBatchDataset operation.static DeserializeIteratorDeserializeIterator.create(Scope scope, Operand<? extends TType> resourceHandle, Operand<? extends TType> serialized) Factory method to create a class wrapping a new DeserializeIterator operation.static DirectedInterleaveDatasetDirectedInterleaveDataset.create(Scope scope, Operand<? extends TType> selectorInputDataset, Iterable<Operand<? extends TType>> dataInputDatasets, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, DirectedInterleaveDataset.Options... options) Factory method to create a class wrapping a new DirectedInterleaveDataset operation.static FilterByLastComponentDatasetFilterByLastComponentDataset.create(Scope scope, Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new FilterByLastComponentDataset operation.static FilterDatasetFilterDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction predicate, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, FilterDataset.Options... options) Factory method to create a class wrapping a new FilterDataset operation.static FinalizeDatasetFinalizeDataset.create(Scope scope, Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, FinalizeDataset.Options... options) Factory method to create a class wrapping a new FinalizeDataset operation.static FixedLengthRecordDatasetFixedLengthRecordDataset.create(Scope scope, Operand<TString> filenames, Operand<TInt64> headerBytes, Operand<TInt64> recordBytes, Operand<TInt64> footerBytes, Operand<TInt64> bufferSize, Operand<TString> compressionType, FixedLengthRecordDataset.Options... options) Factory method to create a class wrapping a new FixedLengthRecordDatasetV2 operation.static FlatMapDatasetFlatMapDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, FlatMapDataset.Options... options) Factory method to create a class wrapping a new FlatMapDataset operation.static GlobalShuffleDatasetGlobalShuffleDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> seed, Operand<TInt64> seed2, Operand<? extends TType> seedGenerator, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, GlobalShuffleDataset.Options... options) Factory method to create a class wrapping a new GlobalShuffleDataset operation.static GroupByReducerDatasetGroupByReducerDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> keyFuncOtherArguments, Iterable<Operand<?>> initFuncOtherArguments, Iterable<Operand<?>> reduceFuncOtherArguments, Iterable<Operand<?>> finalizeFuncOtherArguments, ConcreteFunction keyFunc, ConcreteFunction initFunc, ConcreteFunction reduceFunc, ConcreteFunction finalizeFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new GroupByReducerDataset operation.static GroupByWindowDatasetGroupByWindowDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> keyFuncOtherArguments, Iterable<Operand<?>> reduceFuncOtherArguments, Iterable<Operand<?>> windowSizeFuncOtherArguments, ConcreteFunction keyFunc, ConcreteFunction reduceFunc, ConcreteFunction windowSizeFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, GroupByWindowDataset.Options... options) Factory method to create a class wrapping a new GroupByWindowDataset operation.static IgnoreErrorsDatasetIgnoreErrorsDataset.create(Scope scope, Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, IgnoreErrorsDataset.Options... options) Factory method to create a class wrapping a new IgnoreErrorsDataset operation.static IndexFlatMapDatasetIndexFlatMapDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> mapFuncOtherArgs, Iterable<Operand<?>> indexMapFuncOtherArgs, Operand<TInt64> outputCardinality, ConcreteFunction mapFunc, ConcreteFunction indexMapFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, IndexFlatMapDataset.Options... options) Factory method to create a class wrapping a new IndexFlatMapDataset operation.static InitializeTableFromDatasetInitializeTableFromDataset.create(Scope scope, Operand<? extends TType> tableHandle, Operand<? extends TType> dataset) Factory method to create a class wrapping a new InitializeTableFromDataset operation.static InterleaveDatasetInterleaveDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> cycleLength, Operand<TInt64> blockLength, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, InterleaveDataset.Options... options) Factory method to create a class wrapping a new InterleaveDataset operation.static IteratorFromStringHandleIteratorFromStringHandle.create(Scope scope, Operand<TString> stringHandle, List<Class<? extends TType>> outputTypes, IteratorFromStringHandle.Options... options) Factory method to create a class wrapping a new IteratorFromStringHandleV2 operation.static IteratorGetDeviceFactory method to create a class wrapping a new IteratorGetDevice operation.static IteratorGetModelProtoFactory method to create a class wrapping a new IteratorGetModelProto operation.static IteratorGetNextIteratorGetNext.create(Scope scope, Operand<? extends TType> iterator, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new IteratorGetNext operation.static IteratorGetNextAsOptionalIteratorGetNextAsOptional.create(Scope scope, Operand<? extends TType> iterator, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new IteratorGetNextAsOptional operation.static IteratorGetNextSyncIteratorGetNextSync.create(Scope scope, Operand<? extends TType> iterator, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new IteratorGetNextSync operation.static IteratorToStringHandleFactory method to create a class wrapping a new IteratorToStringHandle operation.static LatencyStatsDatasetLatencyStatsDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TString> tag, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new LatencyStatsDataset operation.static <T extends TNumber>
LeakyReluGrad<T> LeakyReluGrad.create(Scope scope, Operand<T> gradients, Operand<T> features, LeakyReluGrad.Options... options) Factory method to create a class wrapping a new LeakyReluGrad operation.LegacyParallelInterleaveDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> cycleLength, Operand<TInt64> blockLength, Operand<TInt64> bufferOutputElements, Operand<TInt64> prefetchInputElements, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, LegacyParallelInterleaveDataset.Options... options) Factory method to create a class wrapping a new LegacyParallelInterleaveDatasetV2 operation.static ListSnapshotChunksDatasetListSnapshotChunksDataset.create(Scope scope, Operand<TString> snapshotPath, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ListSnapshotChunksDataset operation.static LMDBDatasetLMDBDataset.create(Scope scope, Operand<TString> filenames, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new LMDBDataset operation.static LoadDatasetLoadDataset.create(Scope scope, Operand<TString> path, Iterable<Operand<?>> readerFuncOtherArgs, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ConcreteFunction readerFunc, LoadDataset.Options... options) Factory method to create a class wrapping a new LoadDataset operation.static MakeIteratorMakeIterator.create(Scope scope, Operand<? extends TType> dataset, Operand<? extends TType> iterator) Factory method to create a class wrapping a new MakeIterator operation.static MapAndBatchDatasetMapAndBatchDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> batchSize, Operand<TInt64> numParallelCalls, Operand<TBool> dropRemainder, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, MapAndBatchDataset.Options... options) Factory method to create a class wrapping a new MapAndBatchDataset operation.static MapDatasetMapDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, MapDataset.Options... options) Factory method to create a class wrapping a new MapDataset operation.static MatchingFilesDatasetFactory method to create a class wrapping a new MatchingFilesDataset operation.static MaxIntraOpParallelismDatasetMaxIntraOpParallelismDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> maxIntraOpParallelism, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new MaxIntraOpParallelismDataset operation.static ModelDatasetModelDataset.create(Scope scope, Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ModelDataset.Options... options) Factory method to create a class wrapping a new ModelDataset operation.MultiDeviceIteratorFromStringHandle.create(Scope scope, Operand<TString> stringHandle, List<Class<? extends TType>> outputTypes, MultiDeviceIteratorFromStringHandle.Options... options) Factory method to create a class wrapping a new MultiDeviceIteratorFromStringHandle operation.MultiDeviceIteratorGetNextFromShard.create(Scope scope, Operand<? extends TType> multiDeviceIterator, Operand<TInt32> shardNum, Operand<TInt64> incarnationId, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new MultiDeviceIteratorGetNextFromShard operation.static MultiDeviceIteratorInitMultiDeviceIteratorInit.create(Scope scope, Operand<? extends TType> dataset, Operand<? extends TType> multiDeviceIterator, Operand<TInt64> maxBufferSize) Factory method to create a class wrapping a new MultiDeviceIteratorInit operation.Factory method to create a class wrapping a new MultiDeviceIteratorToStringHandle operation.static NonSerializableDatasetNonSerializableDataset.create(Scope scope, Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new NonSerializableDataset operation.static OptimizeDatasetOptimizeDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TString> optimizationsEnabled, Operand<TString> optimizationsDisabled, Operand<TString> optimizationsDefault, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, OptimizeDataset.Options... options) Factory method to create a class wrapping a new OptimizeDatasetV2 operation.static OptionalGetValueOptionalGetValue.create(Scope scope, Operand<? extends TType> optional, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new OptionalGetValue operation.static OptionalHasValueFactory method to create a class wrapping a new OptionalHasValue operation.static OptionsDatasetOptionsDataset.create(Scope scope, Operand<? extends TType> inputDataset, String serializedOptions, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, OptionsDataset.Options... options) Factory method to create a class wrapping a new OptionsDataset operation.static PaddedBatchDatasetPaddedBatchDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> batchSize, Iterable<Operand<TInt64>> paddedShapes, Iterable<Operand<?>> paddingValues, Operand<TBool> dropRemainder, List<Shape> outputShapes, PaddedBatchDataset.Options... options) Factory method to create a class wrapping a new PaddedBatchDatasetV2 operation.static ParallelBatchDatasetParallelBatchDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> batchSize, Operand<TInt64> numParallelCalls, Operand<TBool> dropRemainder, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ParallelBatchDataset.Options... options) Factory method to create a class wrapping a new ParallelBatchDataset operation.static ParallelFilterDatasetParallelFilterDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> numParallelCalls, ConcreteFunction predicate, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ParallelFilterDataset.Options... options) Factory method to create a class wrapping a new ParallelFilterDataset operation.static ParallelInterleaveDatasetParallelInterleaveDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> cycleLength, Operand<TInt64> blockLength, Operand<TInt64> bufferOutputElements, Operand<TInt64> prefetchInputElements, Operand<TInt64> numParallelCalls, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ParallelInterleaveDataset.Options... options) Factory method to create a class wrapping a new ParallelInterleaveDatasetV4 operation.static ParallelMapDatasetParallelMapDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> numParallelCalls, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ParallelMapDataset.Options... options) Factory method to create a class wrapping a new ParallelMapDatasetV2 operation.static ParseExampleDatasetParseExampleDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> numParallelCalls, Iterable<Operand<?>> denseDefaults, List<String> sparseKeys, List<String> denseKeys, List<Class<? extends TType>> sparseTypes, List<Shape> denseShapes, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, List<Class<? extends TType>> raggedValueTypes, List<Class<? extends TNumber>> raggedSplitTypes, ParseExampleDataset.Options... options) Factory method to create a class wrapping a new ParseExampleDatasetV2 operation.static PrefetchDatasetPrefetchDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> bufferSize, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, PrefetchDataset.Options... options) Factory method to create a class wrapping a new PrefetchDataset operation.static PrivateThreadPoolDatasetPrivateThreadPoolDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> numThreads, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new PrivateThreadPoolDataset operation.static RandomDatasetRandomDataset.create(Scope scope, Operand<TInt64> seed, Operand<TInt64> seed2, Operand<? extends TType> seedGenerator, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, RandomDataset.Options... options) Factory method to create a class wrapping a new RandomDatasetV2 operation.static RangeDatasetRangeDataset.create(Scope scope, Operand<TInt64> start, Operand<TInt64> stop, Operand<TInt64> step, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, RangeDataset.Options... options) Factory method to create a class wrapping a new RangeDataset operation.static RebatchDatasetV2RebatchDatasetV2.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> batchSizes, Operand<TBool> dropRemainder, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new RebatchDatasetV2 operation.static ReduceDatasetReduceDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> initialState, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ReduceDataset.Options... options) Factory method to create a class wrapping a new ReduceDataset operation.static RegisterDatasetRegisterDataset.create(Scope scope, Operand<? extends TType> dataset, Operand<TString> address, Operand<TString> protocol, Long externalStatePolicy, RegisterDataset.Options... options) Factory method to create a class wrapping a new RegisterDatasetV2 operation.static RepeatDatasetRepeatDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> count, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, RepeatDataset.Options... options) Factory method to create a class wrapping a new RepeatDataset operation.static RewriteDatasetRewriteDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TString> rewriteName, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new RewriteDataset operation.static SamplingDatasetSamplingDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TFloat32> rate, Operand<TInt64> seed, Operand<TInt64> seed2, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new SamplingDataset operation.static SaveDatasetSaveDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TString> path, Iterable<Operand<?>> shardFuncOtherArgs, ConcreteFunction shardFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, SaveDataset.Options... options) Factory method to create a class wrapping a new SaveDatasetV2 operation.static ScanDatasetScanDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> initialState, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ScanDataset.Options... options) Factory method to create a class wrapping a new ScanDataset operation.static SerializeIteratorSerializeIterator.create(Scope scope, Operand<? extends TType> resourceHandle, SerializeIterator.Options... options) Factory method to create a class wrapping a new SerializeIterator operation.static SetStatsAggregatorDatasetSetStatsAggregatorDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<? extends TType> statsAggregator, Operand<TString> tag, Operand<TString> counterPrefix, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new SetStatsAggregatorDataset operation.static ShardDatasetShardDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> numShards, Operand<TInt64> index, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ShardDataset.Options... options) Factory method to create a class wrapping a new ShardDataset operation.static ShuffleAndRepeatDatasetShuffleAndRepeatDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> bufferSize, Operand<TInt64> seed, Operand<TInt64> seed2, Operand<TInt64> count, Operand<? extends TType> seedGenerator, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ShuffleAndRepeatDataset.Options... options) Factory method to create a class wrapping a new ShuffleAndRepeatDatasetV2 operation.static ShuffleDatasetShuffleDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> bufferSize, Operand<TInt64> seed, Operand<TInt64> seed2, Operand<? extends TType> seedGenerator, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ShuffleDataset.Options... options) Factory method to create a class wrapping a new ShuffleDatasetV3 operation.static SkipDatasetSkipDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> count, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, SkipDataset.Options... options) Factory method to create a class wrapping a new SkipDataset operation.static SleepDatasetSleepDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> sleepMicroseconds, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new SleepDataset operation.static SlidingWindowDatasetSlidingWindowDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> windowSize, Operand<TInt64> windowShift, Operand<TInt64> windowStride, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, SlidingWindowDataset.Options... options) Factory method to create a class wrapping a new SlidingWindowDataset operation.static SnapshotChunkDatasetSnapshotChunkDataset.create(Scope scope, Operand<TString> chunkFile, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, SnapshotChunkDataset.Options... options) Factory method to create a class wrapping a new SnapshotChunkDataset operation.static SnapshotDatasetSnapshotDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TString> path, Iterable<Operand<?>> readerFuncOtherArgs, Iterable<Operand<?>> shardFuncOtherArgs, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ConcreteFunction readerFunc, ConcreteFunction shardFunc, SnapshotDataset.Options... options) Factory method to create a class wrapping a new SnapshotDatasetV2 operation.static SnapshotDatasetReaderSnapshotDatasetReader.create(Scope scope, Operand<TString> shardDir, Operand<TInt64> startIndex, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, Long version, SnapshotDatasetReader.Options... options) Factory method to create a class wrapping a new SnapshotDatasetReader operation.static SparseTensorSliceDatasetSparseTensorSliceDataset.create(Scope scope, Operand<TInt64> indices, Operand<? extends TType> values, Operand<TInt64> denseShape) Factory method to create a class wrapping a new SparseTensorSliceDataset operation.static SqlDatasetSqlDataset.create(Scope scope, Operand<TString> driverName, Operand<TString> dataSourceName, Operand<TString> query, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new SqlDataset operation.StatsAggregatorSetSummaryWriter.create(Scope scope, Operand<? extends TType> statsAggregator, Operand<? extends TType> summary) Factory method to create a class wrapping a new StatsAggregatorSetSummaryWriter operation.static TakeDatasetTakeDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> count, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, TakeDataset.Options... options) Factory method to create a class wrapping a new TakeDataset operation.static TakeWhileDatasetTakeWhileDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction predicate, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, TakeWhileDataset.Options... options) Factory method to create a class wrapping a new TakeWhileDataset operation.static TextLineDatasetTextLineDataset.create(Scope scope, Operand<TString> filenames, Operand<TString> compressionType, Operand<TInt64> bufferSize, TextLineDataset.Options... options) Factory method to create a class wrapping a new TextLineDataset operation.static TfRecordDatasetTfRecordDataset.create(Scope scope, Operand<TString> filenames, Operand<TString> compressionType, Operand<TInt64> bufferSize, Operand<TInt64> byteOffsets, TfRecordDataset.Options... options) Factory method to create a class wrapping a new TFRecordDatasetV2 operation.static ThreadPoolDatasetThreadPoolDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<? extends TType> threadPool, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ThreadPoolDataset operation.static UnbatchDatasetUnbatchDataset.create(Scope scope, Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, UnbatchDataset.Options... options) Factory method to create a class wrapping a new UnbatchDataset operation.static UncompressElementUncompressElement.create(Scope scope, Operand<? extends TType> compressed, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new UncompressElement operation.static UniqueDatasetUniqueDataset.create(Scope scope, Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, UniqueDataset.Options... options) Factory method to create a class wrapping a new UniqueDataset operation.static UnwrapDatasetVariantFactory method to create a class wrapping a new UnwrapDatasetVariant operation.static WindowDatasetWindowDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> sizeOutput, Operand<TInt64> shift, Operand<TInt64> stride, Operand<TBool> dropRemainder, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, WindowDataset.Options... options) Factory method to create a class wrapping a new WindowDataset operation.static WrapDatasetVariantFactory method to create a class wrapping a new WrapDatasetVariant operation.Method parameters in org.tensorflow.op.data with type arguments of type OperandModifier and TypeMethodDescriptionstatic ChooseFastestBranchDatasetChooseFastestBranchDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> ratioNumerator, Operand<TInt64> ratioDenominator, Iterable<Operand<?>> otherArguments, Long numElementsPerBranch, List<ConcreteFunction> branches, List<Long> otherArgumentsLengths, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ChooseFastestBranchDataset operation.static ChooseFastestDatasetChooseFastestDataset.create(Scope scope, Iterable<Operand<? extends TType>> inputDatasets, Long numExperiments, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ChooseFastestDataset operation.static CompressElementFactory method to create a class wrapping a new CompressElement operation.static CSVDatasetCSVDataset.create(Scope scope, Operand<TString> filenames, Operand<TString> compressionType, Operand<TInt64> bufferSize, Operand<TBool> header, Operand<TString> fieldDelim, Operand<TBool> useQuoteDelim, Operand<TString> naValue, Operand<TInt64> selectCols, Iterable<Operand<?>> recordDefaults, Operand<TInt64> excludeCols, List<Shape> outputShapes) Factory method to create a class wrapping a new CSVDatasetV2 operation.static DeleteMultiDeviceIteratorDeleteMultiDeviceIterator.create(Scope scope, Operand<? extends TType> multiDeviceIterator, Iterable<Operand<? extends TType>> iterators, Operand<? extends TType> deleter) Factory method to create a class wrapping a new DeleteMultiDeviceIterator operation.static DirectedInterleaveDatasetDirectedInterleaveDataset.create(Scope scope, Operand<? extends TType> selectorInputDataset, Iterable<Operand<? extends TType>> dataInputDatasets, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, DirectedInterleaveDataset.Options... options) Factory method to create a class wrapping a new DirectedInterleaveDataset operation.static FilterDatasetFilterDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction predicate, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, FilterDataset.Options... options) Factory method to create a class wrapping a new FilterDataset operation.static FlatMapDatasetFlatMapDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, FlatMapDataset.Options... options) Factory method to create a class wrapping a new FlatMapDataset operation.static GeneratorDatasetGeneratorDataset.create(Scope scope, Iterable<Operand<?>> initFuncOtherArgs, Iterable<Operand<?>> nextFuncOtherArgs, Iterable<Operand<?>> finalizeFuncOtherArgs, ConcreteFunction initFunc, ConcreteFunction nextFunc, ConcreteFunction finalizeFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, GeneratorDataset.Options... options) Factory method to create a class wrapping a new GeneratorDataset operation.static GroupByReducerDatasetGroupByReducerDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> keyFuncOtherArguments, Iterable<Operand<?>> initFuncOtherArguments, Iterable<Operand<?>> reduceFuncOtherArguments, Iterable<Operand<?>> finalizeFuncOtherArguments, ConcreteFunction keyFunc, ConcreteFunction initFunc, ConcreteFunction reduceFunc, ConcreteFunction finalizeFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new GroupByReducerDataset operation.static GroupByWindowDatasetGroupByWindowDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> keyFuncOtherArguments, Iterable<Operand<?>> reduceFuncOtherArguments, Iterable<Operand<?>> windowSizeFuncOtherArguments, ConcreteFunction keyFunc, ConcreteFunction reduceFunc, ConcreteFunction windowSizeFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, GroupByWindowDataset.Options... options) Factory method to create a class wrapping a new GroupByWindowDataset operation.static IndexFlatMapDatasetIndexFlatMapDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> mapFuncOtherArgs, Iterable<Operand<?>> indexMapFuncOtherArgs, Operand<TInt64> outputCardinality, ConcreteFunction mapFunc, ConcreteFunction indexMapFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, IndexFlatMapDataset.Options... options) Factory method to create a class wrapping a new IndexFlatMapDataset operation.static InterleaveDatasetInterleaveDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> cycleLength, Operand<TInt64> blockLength, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, InterleaveDataset.Options... options) Factory method to create a class wrapping a new InterleaveDataset operation.LegacyParallelInterleaveDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> cycleLength, Operand<TInt64> blockLength, Operand<TInt64> bufferOutputElements, Operand<TInt64> prefetchInputElements, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, LegacyParallelInterleaveDataset.Options... options) Factory method to create a class wrapping a new LegacyParallelInterleaveDatasetV2 operation.static ListDatasetListDataset.create(Scope scope, Iterable<Operand<?>> tensors, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ListDataset.Options... options) Factory method to create a class wrapping a new ListDataset operation.static LoadDatasetLoadDataset.create(Scope scope, Operand<TString> path, Iterable<Operand<?>> readerFuncOtherArgs, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ConcreteFunction readerFunc, LoadDataset.Options... options) Factory method to create a class wrapping a new LoadDataset operation.static MapAndBatchDatasetMapAndBatchDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> batchSize, Operand<TInt64> numParallelCalls, Operand<TBool> dropRemainder, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, MapAndBatchDataset.Options... options) Factory method to create a class wrapping a new MapAndBatchDataset operation.static MapDatasetMapDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, MapDataset.Options... options) Factory method to create a class wrapping a new MapDataset operation.static OptionalFromValueFactory method to create a class wrapping a new OptionalFromValue operation.static PaddedBatchDatasetPaddedBatchDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> batchSize, Iterable<Operand<TInt64>> paddedShapes, Iterable<Operand<?>> paddingValues, Operand<TBool> dropRemainder, List<Shape> outputShapes, PaddedBatchDataset.Options... options) Factory method to create a class wrapping a new PaddedBatchDatasetV2 operation.static ParallelFilterDatasetParallelFilterDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> numParallelCalls, ConcreteFunction predicate, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ParallelFilterDataset.Options... options) Factory method to create a class wrapping a new ParallelFilterDataset operation.static ParallelInterleaveDatasetParallelInterleaveDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> cycleLength, Operand<TInt64> blockLength, Operand<TInt64> bufferOutputElements, Operand<TInt64> prefetchInputElements, Operand<TInt64> numParallelCalls, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ParallelInterleaveDataset.Options... options) Factory method to create a class wrapping a new ParallelInterleaveDatasetV4 operation.static ParallelMapDatasetParallelMapDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> numParallelCalls, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ParallelMapDataset.Options... options) Factory method to create a class wrapping a new ParallelMapDatasetV2 operation.static ParseExampleDatasetParseExampleDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> numParallelCalls, Iterable<Operand<?>> denseDefaults, List<String> sparseKeys, List<String> denseKeys, List<Class<? extends TType>> sparseTypes, List<Shape> denseShapes, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, List<Class<? extends TType>> raggedValueTypes, List<Class<? extends TNumber>> raggedSplitTypes, ParseExampleDataset.Options... options) Factory method to create a class wrapping a new ParseExampleDatasetV2 operation.static ReduceDatasetReduceDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> initialState, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ReduceDataset.Options... options) Factory method to create a class wrapping a new ReduceDataset operation.static SaveDatasetSaveDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TString> path, Iterable<Operand<?>> shardFuncOtherArgs, ConcreteFunction shardFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, SaveDataset.Options... options) Factory method to create a class wrapping a new SaveDatasetV2 operation.static ScanDatasetScanDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> initialState, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ScanDataset.Options... options) Factory method to create a class wrapping a new ScanDataset operation.static SnapshotDatasetSnapshotDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TString> path, Iterable<Operand<?>> readerFuncOtherArgs, Iterable<Operand<?>> shardFuncOtherArgs, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ConcreteFunction readerFunc, ConcreteFunction shardFunc, SnapshotDataset.Options... options) Factory method to create a class wrapping a new SnapshotDatasetV2 operation.static SnapshotNestedDatasetReaderSnapshotNestedDatasetReader.create(Scope scope, Iterable<Operand<? extends TType>> inputs, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new SnapshotNestedDatasetReader operation.static TakeWhileDatasetTakeWhileDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction predicate, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, TakeWhileDataset.Options... options) Factory method to create a class wrapping a new TakeWhileDataset operation.static TensorDatasetTensorDataset.create(Scope scope, Iterable<Operand<?>> components, List<Shape> outputShapes, TensorDataset.Options... options) Factory method to create a class wrapping a new TensorDataset operation.static TensorSliceDatasetTensorSliceDataset.create(Scope scope, Iterable<Operand<?>> components, List<Shape> outputShapes, TensorSliceDataset.Options... options) Factory method to create a class wrapping a new TensorSliceDataset operation.static WeightedFlatMapDatasetWeightedFlatMapDataset.create(Scope scope, Iterable<Operand<? extends TType>> inputDatasets, Iterable<Operand<TFloat64>> weights, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, WeightedFlatMapDataset.Options... options) Factory method to create a class wrapping a new WeightedFlatMapDataset operation.static WindowOpWindowOp.create(Scope scope, Iterable<Operand<?>> inputs, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new WindowOp operation.static ZipDatasetZipDataset.create(Scope scope, Iterable<Operand<? extends TType>> inputDatasets, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ZipDataset.Options... options) Factory method to create a class wrapping a new ZipDataset operation. -
Uses of Operand in org.tensorflow.op.data.experimental
Classes in org.tensorflow.op.data.experimental that implement OperandModifier and TypeClassDescriptionfinal classThe ExperimentalAssertNextDataset operationfinal classCreates a dataset that shards the input dataset.final classRecords the bytes size of each element ofinput_datasetin a StatsAggregator.final classThe ExperimentalChooseFastestDataset operationfinal classThe ExperimentalCSVDataset operationfinal classReturns the cardinality ofinput_dataset.final classCreates a dataset that batches input elements into a SparseTensor.final classA substitute forInterleaveDataseton a fixed list ofNdatasets.final classCreates a dataset that computes a group-by oninput_dataset.final classCreates a dataset that computes a windowed group-by oninput_dataset. // TODO(mrry): Support non-int64 keys.final classCreates a dataset that contains the elements ofinput_datasetignoring errors.final classReturns the name of the device on whichresourcehas been placed.final classRecords the latency of producinginput_datasetelements in a StatsAggregator.final classThe ExperimentalLMDBDataset operationfinal classCreates a dataset that fuses mapping with batching.final classCreates a dataset that appliesfto the outputs ofinput_dataset.final classThe ExperimentalMatchingFilesDataset operationfinal classCreates a dataset that overrides the maximum intra-op parallelism.final classThe ExperimentalNonSerializableDataset operationfinal classCreates a dataset that appliesfto the outputs ofinput_dataset.final classTransformsinput_datasetcontainingExampleprotos as vectors of DT_STRING into a dataset ofTensororSparseTensorobjects representing the parsed features.final classCreates a dataset that uses a custom thread pool to computeinput_dataset.final classCreates a Dataset that returns pseudorandom numbers.final classCreates a dataset that changes the batch size.final classCreates a dataset successively reducesfover the elements ofinput_dataset.final classThe ExperimentalSetStatsAggregatorDataset operationfinal classThe ExperimentalSleepDataset operationfinal classCreates a dataset that passes a sliding window overinput_dataset.final classCreates a dataset that executes a SQL query and emits rows of the result set.final classCreates a statistics manager resource.final classProduces a summary of any statistics recorded by the given statistics manager.final classCreates a dataset that stops iteration when predicate` is false.final classCreates a dataset that uses a custom thread pool to computeinput_dataset.final classCreates a dataset that uses a custom thread pool to computeinput_dataset.final classA dataset that splits the elements of its input into multiple elements.final classCreates a dataset that contains the unique elements ofinput_dataset.Fields in org.tensorflow.op.data.experimental declared as OperandModifier and TypeFieldDescriptionDenseToSparseBatchDataset.Inputs.batchSizeA scalar representing the number of elements to accumulate in a batch.MapAndBatchDataset.Inputs.batchSizeA scalar representing the number of elements to accumulate in a batch.ParallelInterleaveDataset.Inputs.blockLengthThe blockLength inputParallelInterleaveDataset.Inputs.bufferOutputElementsThe bufferOutputElements inputCSVDataset.Inputs.bufferSizeThe bufferSize inputCSVDataset.Inputs.compressionTypeThe compressionType inputDatasetToTFRecord.Inputs.compressionTypeA scalar string tensor containing either (i) the empty string (no compression), (ii) "ZLIB", or (iii) "GZIP".SetStatsAggregatorDataset.Inputs.counterPrefixThe counterPrefix inputParallelInterleaveDataset.Inputs.cycleLengthThe cycleLength inputSqlDataset.Inputs.dataSourceNameA connection string to connect to the database.SqlDataset.Inputs.driverNameThe database type.MapAndBatchDataset.Inputs.dropRemainderA scalar representing whether the last batch should be dropped in case its size is smaller than desired.CSVDataset.Inputs.fieldDelimThe fieldDelim inputDatasetToTFRecord.Inputs.filenameA scalar string tensor representing the filename to use.CSVDataset.Inputs.filenamesThe filenames inputLmdbDataset.Inputs.filenamesThe filenames inputCSVDataset.Inputs.headerThe header inputAutoShardDataset.Inputs.indexA scalar representing the index of the current worker out of num_workers.AssertNextDataset.Inputs.inputDatasetThe inputDataset inputAutoShardDataset.Inputs.inputDatasetA variant tensor representing the input dataset.BytesProducedStatsDataset.Inputs.inputDatasetThe inputDataset inputDatasetCardinality.Inputs.inputDatasetA variant tensor representing the dataset to return cardinality for.DatasetToTFRecord.Inputs.inputDatasetA variant tensor representing the dataset to write.DenseToSparseBatchDataset.Inputs.inputDatasetA handle to an input dataset.GroupByReducerDataset.Inputs.inputDatasetA variant tensor representing the input dataset.GroupByWindowDataset.Inputs.inputDatasetThe inputDataset inputIgnoreErrorsDataset.Inputs.inputDatasetThe inputDataset inputLatencyStatsDataset.Inputs.inputDatasetThe inputDataset inputMapAndBatchDataset.Inputs.inputDatasetA variant tensor representing the input dataset.MapDataset.Inputs.inputDatasetThe inputDataset inputMaxIntraOpParallelismDataset.Inputs.inputDatasetThe inputDataset inputNonSerializableDataset.Inputs.inputDatasetThe inputDataset inputParallelInterleaveDataset.Inputs.inputDatasetThe inputDataset inputParseExampleDataset.Inputs.inputDatasetThe inputDataset inputPrivateThreadPoolDataset.Inputs.inputDatasetThe inputDataset inputRebatchDataset.Inputs.inputDatasetA variant tensor representing the input dataset.ScanDataset.Inputs.inputDatasetThe inputDataset inputSetStatsAggregatorDataset.Inputs.inputDatasetThe inputDataset inputSleepDataset.Inputs.inputDatasetThe inputDataset inputSlidingWindowDataset.Inputs.inputDatasetThe inputDataset inputTakeWhileDataset.Inputs.inputDatasetThe inputDataset inputThreadPoolDataset.Inputs.inputDatasetThe inputDataset inputUnbatchDataset.Inputs.inputDatasetThe inputDataset inputUniqueDataset.Inputs.inputDatasetThe inputDataset inputStatsAggregatorSummary.Inputs.iteratorThe iterator inputMaxIntraOpParallelismDataset.Inputs.maxIntraOpParallelismIdentifies the maximum intra-op parallelism to use.CSVDataset.Inputs.naValueThe naValue inputMapAndBatchDataset.Inputs.numParallelCallsA scalar representing the maximum number of parallel invocations of themap_fnfunction.ParseExampleDataset.Inputs.numParallelCallsThe numParallelCalls inputRebatchDataset.Inputs.numReplicasA scalar representing the number of replicas to distribute this batch across.PrivateThreadPoolDataset.Inputs.numThreadsIdentifies the number of threads to use for the private threadpool.AutoShardDataset.Inputs.numWorkersA scalar representing the number of workers to distribute this dataset across.MatchingFilesDataset.Inputs.patternsThe patterns inputParallelInterleaveDataset.Inputs.prefetchInputElementsThe prefetchInputElements inputSqlDataset.Inputs.queryA SQL query to execute.IteratorGetDevice.Inputs.resourceThe resource inputDenseToSparseBatchDataset.Inputs.rowShapeA vector representing the dense shape of each row in the produced SparseTensor.RandomDataset.Inputs.seedA scalar seed for the random number generator.RandomDataset.Inputs.seed2A second scalar seed to avoid seed collision.CSVDataset.Inputs.selectColsThe selectCols inputDirectedInterleaveDataset.Inputs.selectorInputDatasetA dataset of scalarDT_INT64elements that determines which of theNdata inputs should produce the next output element.SleepDataset.Inputs.sleepMicrosecondsThe sleepMicroseconds inputParallelInterleaveDataset.Inputs.sloppyThe sloppy inputSetStatsAggregatorDataset.Inputs.statsAggregatorThe statsAggregator inputBytesProducedStatsDataset.Inputs.tagThe tag inputLatencyStatsDataset.Inputs.tagThe tag inputSetStatsAggregatorDataset.Inputs.tagThe tag inputThreadPoolDataset.Inputs.threadPoolA resource produced by the ThreadPoolHandle op.AssertNextDataset.Inputs.transformationsThe transformations inputCSVDataset.Inputs.useQuoteDelimThe useQuoteDelim inputSlidingWindowDataset.Inputs.windowShiftA scalar representing the steps moving the sliding window forward in one iteration.SlidingWindowDataset.Inputs.windowSizeA scalar representing the number of elements in the sliding window.SlidingWindowDataset.Inputs.windowStrideA scalar representing the stride of the input elements of the sliding window.Fields in org.tensorflow.op.data.experimental with type parameters of type OperandModifier and TypeFieldDescriptionDirectedInterleaveDataset.Inputs.dataInputDatasetsNdatasets with the same type that will be interleaved according to the values ofselector_input_dataset.ParseExampleDataset.Inputs.denseDefaultsA dict mapping string keys toTensors.GroupByReducerDataset.Inputs.finalizeFuncOtherArgumentsA list of tensors, typically values that were captured when building a closure forfinalize_func.GroupByReducerDataset.Inputs.initFuncOtherArgumentsA list of tensors, typically values that were captured when building a closure forinit_func.ScanDataset.Inputs.initialStateThe initialState inputChooseFastestDataset.Inputs.inputDatasetsThe inputDatasets inputGroupByReducerDataset.Inputs.keyFuncOtherArgumentsA list of tensors, typically values that were captured when building a closure forkey_func.GroupByWindowDataset.Inputs.keyFuncOtherArgumentsThe keyFuncOtherArguments inputMapAndBatchDataset.Inputs.otherArgumentsA list of tensors, typically values that were captured when building a closure forf.MapDataset.Inputs.otherArgumentsThe otherArguments inputParallelInterleaveDataset.Inputs.otherArgumentsThe otherArguments inputScanDataset.Inputs.otherArgumentsThe otherArguments inputTakeWhileDataset.Inputs.otherArgumentsA list of tensors, typically values that were captured when building a closure forpredicate.CSVDataset.Inputs.recordDefaultsThe recordDefaults inputGroupByReducerDataset.Inputs.reduceFuncOtherArgumentsA list of tensors, typically values that were captured when building a closure forreduce_func.GroupByWindowDataset.Inputs.reduceFuncOtherArgumentsThe reduceFuncOtherArguments inputGroupByWindowDataset.Inputs.windowSizeFuncOtherArgumentsThe windowSizeFuncOtherArguments inputMethods in org.tensorflow.op.data.experimental with parameters of type OperandModifier and TypeMethodDescriptionstatic AssertNextDatasetAssertNextDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TString> transformations, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalAssertNextDataset operation.static AutoShardDatasetAutoShardDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> numWorkers, Operand<TInt64> index, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, AutoShardDataset.Options... options) Factory method to create a class wrapping a new ExperimentalAutoShardDataset operation.static BytesProducedStatsDatasetBytesProducedStatsDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TString> tag, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalBytesProducedStatsDataset operation.static CSVDatasetCSVDataset.create(Scope scope, Operand<TString> filenames, Operand<TString> compressionType, Operand<TInt64> bufferSize, Operand<TBool> header, Operand<TString> fieldDelim, Operand<TBool> useQuoteDelim, Operand<TString> naValue, Operand<TInt64> selectCols, Iterable<Operand<?>> recordDefaults, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalCSVDataset operation.static DatasetCardinalityFactory method to create a class wrapping a new ExperimentalDatasetCardinality operation.static DatasetToTFRecordDatasetToTFRecord.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TString> filename, Operand<TString> compressionType) Factory method to create a class wrapping a new ExperimentalDatasetToTFRecord operation.static DenseToSparseBatchDatasetDenseToSparseBatchDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> batchSize, Operand<TInt64> rowShape, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalDenseToSparseBatchDataset operation.static DirectedInterleaveDatasetDirectedInterleaveDataset.create(Scope scope, Operand<? extends TType> selectorInputDataset, Iterable<Operand<? extends TType>> dataInputDatasets, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalDirectedInterleaveDataset operation.static GroupByReducerDatasetGroupByReducerDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> keyFuncOtherArguments, Iterable<Operand<?>> initFuncOtherArguments, Iterable<Operand<?>> reduceFuncOtherArguments, Iterable<Operand<?>> finalizeFuncOtherArguments, ConcreteFunction keyFunc, ConcreteFunction initFunc, ConcreteFunction reduceFunc, ConcreteFunction finalizeFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalGroupByReducerDataset operation.static GroupByWindowDatasetGroupByWindowDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> keyFuncOtherArguments, Iterable<Operand<?>> reduceFuncOtherArguments, Iterable<Operand<?>> windowSizeFuncOtherArguments, ConcreteFunction keyFunc, ConcreteFunction reduceFunc, ConcreteFunction windowSizeFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalGroupByWindowDataset operation.static IgnoreErrorsDatasetIgnoreErrorsDataset.create(Scope scope, Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, IgnoreErrorsDataset.Options... options) Factory method to create a class wrapping a new ExperimentalIgnoreErrorsDataset operation.static IteratorGetDeviceFactory method to create a class wrapping a new ExperimentalIteratorGetDevice operation.static LatencyStatsDatasetLatencyStatsDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TString> tag, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalLatencyStatsDataset operation.static LmdbDatasetLmdbDataset.create(Scope scope, Operand<TString> filenames, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalLMDBDataset operation.static MapAndBatchDatasetMapAndBatchDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> batchSize, Operand<TInt64> numParallelCalls, Operand<TBool> dropRemainder, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, MapAndBatchDataset.Options... options) Factory method to create a class wrapping a new ExperimentalMapAndBatchDataset operation.static MapDatasetMapDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, MapDataset.Options... options) Factory method to create a class wrapping a new ExperimentalMapDataset operation.static MatchingFilesDatasetFactory method to create a class wrapping a new ExperimentalMatchingFilesDataset operation.static MaxIntraOpParallelismDatasetMaxIntraOpParallelismDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> maxIntraOpParallelism, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalMaxIntraOpParallelismDataset operation.static NonSerializableDatasetNonSerializableDataset.create(Scope scope, Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalNonSerializableDataset operation.static ParallelInterleaveDatasetParallelInterleaveDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> cycleLength, Operand<TInt64> blockLength, Operand<TBool> sloppy, Operand<TInt64> bufferOutputElements, Operand<TInt64> prefetchInputElements, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalParallelInterleaveDataset operation.static ParseExampleDatasetParseExampleDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> numParallelCalls, Iterable<Operand<?>> denseDefaults, List<String> sparseKeys, List<String> denseKeys, List<Class<? extends TType>> sparseTypes, List<Shape> denseShapes, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ParseExampleDataset.Options... options) Factory method to create a class wrapping a new ExperimentalParseExampleDataset operation.static PrivateThreadPoolDatasetPrivateThreadPoolDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> numThreads, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalPrivateThreadPoolDataset operation.static RandomDatasetRandomDataset.create(Scope scope, Operand<TInt64> seed, Operand<TInt64> seed2, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalRandomDataset operation.static RebatchDatasetRebatchDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> numReplicas, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, RebatchDataset.Options... options) Factory method to create a class wrapping a new ExperimentalRebatchDataset operation.static ScanDatasetScanDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> initialState, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ScanDataset.Options... options) Factory method to create a class wrapping a new ExperimentalScanDataset operation.static SetStatsAggregatorDatasetSetStatsAggregatorDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<? extends TType> statsAggregator, Operand<TString> tag, Operand<TString> counterPrefix, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalSetStatsAggregatorDataset operation.static SleepDatasetSleepDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> sleepMicroseconds, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalSleepDataset operation.static SlidingWindowDatasetSlidingWindowDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> windowSize, Operand<TInt64> windowShift, Operand<TInt64> windowStride, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalSlidingWindowDataset operation.static SqlDatasetSqlDataset.create(Scope scope, Operand<TString> driverName, Operand<TString> dataSourceName, Operand<TString> query, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalSqlDataset operation.static StatsAggregatorSummaryFactory method to create a class wrapping a new ExperimentalStatsAggregatorSummary operation.static TakeWhileDatasetTakeWhileDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction predicate, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalTakeWhileDataset operation.static ThreadPoolDatasetThreadPoolDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<? extends TType> threadPool, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalThreadPoolDataset operation.static UnbatchDatasetUnbatchDataset.create(Scope scope, Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalUnbatchDataset operation.static UniqueDatasetUniqueDataset.create(Scope scope, Operand<? extends TType> inputDataset, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalUniqueDataset operation.Method parameters in org.tensorflow.op.data.experimental with type arguments of type OperandModifier and TypeMethodDescriptionstatic ChooseFastestDatasetChooseFastestDataset.create(Scope scope, Iterable<Operand<? extends TType>> inputDatasets, Long numExperiments, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalChooseFastestDataset operation.static CSVDatasetCSVDataset.create(Scope scope, Operand<TString> filenames, Operand<TString> compressionType, Operand<TInt64> bufferSize, Operand<TBool> header, Operand<TString> fieldDelim, Operand<TBool> useQuoteDelim, Operand<TString> naValue, Operand<TInt64> selectCols, Iterable<Operand<?>> recordDefaults, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalCSVDataset operation.static DirectedInterleaveDatasetDirectedInterleaveDataset.create(Scope scope, Operand<? extends TType> selectorInputDataset, Iterable<Operand<? extends TType>> dataInputDatasets, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalDirectedInterleaveDataset operation.static GroupByReducerDatasetGroupByReducerDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> keyFuncOtherArguments, Iterable<Operand<?>> initFuncOtherArguments, Iterable<Operand<?>> reduceFuncOtherArguments, Iterable<Operand<?>> finalizeFuncOtherArguments, ConcreteFunction keyFunc, ConcreteFunction initFunc, ConcreteFunction reduceFunc, ConcreteFunction finalizeFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalGroupByReducerDataset operation.static GroupByWindowDatasetGroupByWindowDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> keyFuncOtherArguments, Iterable<Operand<?>> reduceFuncOtherArguments, Iterable<Operand<?>> windowSizeFuncOtherArguments, ConcreteFunction keyFunc, ConcreteFunction reduceFunc, ConcreteFunction windowSizeFunc, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalGroupByWindowDataset operation.static MapAndBatchDatasetMapAndBatchDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> batchSize, Operand<TInt64> numParallelCalls, Operand<TBool> dropRemainder, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, MapAndBatchDataset.Options... options) Factory method to create a class wrapping a new ExperimentalMapAndBatchDataset operation.static MapDatasetMapDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, MapDataset.Options... options) Factory method to create a class wrapping a new ExperimentalMapDataset operation.static ParallelInterleaveDatasetParallelInterleaveDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, Operand<TInt64> cycleLength, Operand<TInt64> blockLength, Operand<TBool> sloppy, Operand<TInt64> bufferOutputElements, Operand<TInt64> prefetchInputElements, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalParallelInterleaveDataset operation.static ParseExampleDatasetParseExampleDataset.create(Scope scope, Operand<? extends TType> inputDataset, Operand<TInt64> numParallelCalls, Iterable<Operand<?>> denseDefaults, List<String> sparseKeys, List<String> denseKeys, List<Class<? extends TType>> sparseTypes, List<Shape> denseShapes, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ParseExampleDataset.Options... options) Factory method to create a class wrapping a new ExperimentalParseExampleDataset operation.static ScanDatasetScanDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> initialState, Iterable<Operand<?>> otherArguments, ConcreteFunction f, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes, ScanDataset.Options... options) Factory method to create a class wrapping a new ExperimentalScanDataset operation.static TakeWhileDatasetTakeWhileDataset.create(Scope scope, Operand<? extends TType> inputDataset, Iterable<Operand<?>> otherArguments, ConcreteFunction predicate, List<Class<? extends TType>> outputTypes, List<Shape> outputShapes) Factory method to create a class wrapping a new ExperimentalTakeWhileDataset operation. -
Uses of Operand in org.tensorflow.op.debugging
Classes in org.tensorflow.op.debugging that implement OperandModifier and TypeClassDescriptionfinal classCheckNumerics<T extends TNumber>Checks a tensor for NaN, -Inf and +Inf values.final classDebugGradientIdentity<T extends TType>Identity op for gradient debugging.final classDebugGradientRefIdentity<T extends TType>Identity op for gradient debugging.final classDebugIdentity<T extends TType>Provides an identity mapping of the non-Ref type input tensor for debugging.final classDebug NaN Value Counter Op.final classDebugNumericsSummary<U extends TNumber>Debug Numeric Summary V2 Op.Fields in org.tensorflow.op.debugging declared as OperandModifier and TypeFieldDescriptionDebugGradientIdentity.Inputs.inputThe input inputDebugGradientRefIdentity.Inputs.inputThe input inputDebugIdentity.Inputs.inputInput tensor, non-Reference typeDebugNanCount.Inputs.inputInput tensor, non-Reference type.DebugNumericsSummary.Inputs.inputInput tensor, to be summarized by the op.CheckNumerics.Inputs.tensorThe tensor inputMethods in org.tensorflow.op.debugging with parameters of type OperandModifier and TypeMethodDescriptionstatic <T extends TNumber>
CheckNumerics<T> Factory method to create a class wrapping a new CheckNumericsV2 operation.static <T extends TType>
DebugGradientIdentity<T> Factory method to create a class wrapping a new DebugGradientIdentity operation.static <T extends TType>
DebugGradientRefIdentity<T> Factory method to create a class wrapping a new DebugGradientRefIdentity operation.static <T extends TType>
DebugIdentity<T> DebugIdentity.create(Scope scope, Operand<T> input, DebugIdentity.Options... options) Factory method to create a class wrapping a new DebugIdentityV3 operation.static DebugNanCountDebugNanCount.create(Scope scope, Operand<? extends TType> input, DebugNanCount.Options... options) Factory method to create a class wrapping a new DebugNanCount operation.static <U extends TNumber>
DebugNumericsSummary<U> DebugNumericsSummary.create(Scope scope, Operand<? extends TType> input, Class<U> outputDtype, DebugNumericsSummary.Options... options) Factory method to create a class wrapping a new DebugNumericSummaryV2 operation.static DebugNumericsSummary<TFloat32> DebugNumericsSummary.create(Scope scope, Operand<? extends TType> input, DebugNumericsSummary.Options[] options) Factory method to create a class wrapping a new DebugNumericSummaryV2 operation, with the default output types. -
Uses of Operand in org.tensorflow.op.distribute
Classes in org.tensorflow.op.distribute that implement OperandModifier and TypeClassDescriptionfinal classNcclAllReduce<T extends TNumber>Outputs a tensor containing the reduction across all input tensors.final classNcclBroadcast<T extends TNumber>Sendsinputto all devices that are connected to the output.final classNcclReduce<T extends TNumber>Reducesinputfromnum_devicesusingreductionto a single device.Fields in org.tensorflow.op.distribute declared as OperandModifier and TypeFieldDescriptionNcclAllReduce.Inputs.inputThe input inputNcclBroadcast.Inputs.inputThe input inputFields in org.tensorflow.op.distribute with type parameters of type OperandMethods in org.tensorflow.op.distribute with parameters of type OperandModifier and TypeMethodDescriptionstatic <T extends TNumber>
NcclAllReduce<T> NcclAllReduce.create(Scope scope, Operand<T> input, String reduction, Long numDevices, String sharedName) Factory method to create a class wrapping a new NcclAllReduce operation.static <T extends TNumber>
NcclBroadcast<T> Factory method to create a class wrapping a new NcclBroadcast operation.Method parameters in org.tensorflow.op.distribute with type arguments of type Operand -
Uses of Operand in org.tensorflow.op.dtypes
Classes in org.tensorflow.op.dtypes that implement OperandModifier and TypeClassDescriptionfinal classConverts each entry in the given tensor to strings.final classCast x of type SrcT to y of DstT.final classConverts two real numbers to a complex number.final classConverts a tensor to a scalar predicate.Fields in org.tensorflow.op.dtypes declared as OperandModifier and TypeFieldDescriptionComplex.Inputs.imagThe imag inputAsString.Inputs.inputThe input inputToBool.Inputs.inputThe input inputComplex.Inputs.realThe real inputCast.Inputs.xThe x inputMethods in org.tensorflow.op.dtypes with parameters of type OperandModifier and TypeMethodDescriptionstatic AsStringAsString.create(Scope scope, Operand<? extends TType> input, AsString.Options... options) Factory method to create a class wrapping a new AsString operation.Factory method to create a class wrapping a new Cast operation.Factory method to create a class wrapping a new Complex operation.static ToBoolFactory method to create a class wrapping a new ToBool operation. -
Uses of Operand in org.tensorflow.op.image
Classes in org.tensorflow.op.image that implement OperandModifier and TypeClassDescriptionfinal classAdjustContrast<T extends TNumber>Adjust the contrast of one or more images.final classAdjust the hue of one or more images.final classAdjustSaturation<T extends TNumber>Adjust the saturation of one or more images.final classExtracts crops from the input image tensor and resizes them.final classComputes the gradient of the crop_and_resize op wrt the input boxes tensor.final classCropAndResizeGradImage<T extends TNumber>Computes the gradient of the crop_and_resize op wrt the input image tensor.final classDecode and Crop a JPEG-encoded image to a uint8 tensor.final classDecode the first frame of a BMP-encoded image to a uint8 tensor.final classDecode the frame(s) of a GIF-encoded image to a uint8 tensor.final classDecodeImage<T extends TNumber>Function for decode_bmp, decode_gif, decode_jpeg, and decode_png.final classDecode a JPEG-encoded image to a uint8 tensor.final classDecode a PNG-encoded image to a uint8 or uint16 tensor.final classDrawBoundingBoxes<T extends TNumber>Draw bounding boxes on a batch of images.final classJPEG-encode an image.final classJPEG encode input image with provided compression quality.final classPNG-encode an image.final classExtracts a glimpse from the input tensor.final classExtractImagePatches<T extends TType>Extractpatchesfromimagesand put them in the "depth" output dimension.final classExtractJpegShape<T extends TNumber>Extract the shape information of a JPEG-encoded image.final classConvert one or more images from HSV to RGB.final classImageProjectiveTransformV2<T extends TNumber>Applies the given transform to each of the images.final classImageProjectiveTransformV3<T extends TNumber>Applies the given transform to each of the images.final classGreedily selects a subset of bounding boxes in descending order of score, pruning away boxes that have high overlaps with previously selected boxes.final classRandomCrop<T extends TNumber>Randomly cropimage.final classResizeimagestosizeusing area interpolation.final classResizeimagestosizeusing bicubic interpolation.final classResizeBicubicGrad<T extends TNumber>Computes the gradient of bicubic interpolation.final classResizeimagestosizeusing bilinear interpolation.final classResizeBilinearGrad<T extends TNumber>Computes the gradient of bilinear interpolation.final classResizeNearestNeighbor<T extends TNumber>Resizeimagestosizeusing nearest neighbor interpolation.final classResizeNearestNeighborGrad<T extends TNumber>Computes the gradient of nearest neighbor interpolation.final classConverts one or more images from RGB to HSV.final classThe ScaleAndTranslate operationfinal classScaleAndTranslateGrad<T extends TNumber>The ScaleAndTranslateGrad operationFields in org.tensorflow.op.image declared as OperandModifier and TypeFieldDescriptionGenerateBoundingBoxProposals.Inputs.anchorsA 2-D float tensor of shape[num_anchors, 4]describing the anchor boxes.GenerateBoundingBoxProposals.Inputs.bboxDeltasA 4-D float tensor of shape[num_images, height, width, 4 x num_anchors]. encoding boxes with respec to each anchor.SampleDistortedBoundingBox.Inputs.boundingBoxes3-D with shape[batch, N, 4]describing the N bounding boxes associated with the image.StatelessSampleDistortedBoundingBox.Inputs.boundingBoxes3-D with shape[batch, N, 4]describing the N bounding boxes associated with the image.CombinedNonMaxSuppression.Inputs.boxesA 4-D float tensor of shape[batch_size, num_boxes, q, 4].CropAndResize.Inputs.boxesA 2-D tensor of shape[num_boxes, 4].CropAndResizeGradBoxes.Inputs.boxesA 2-D tensor of shape[num_boxes, 4].CropAndResizeGradImage.Inputs.boxesA 2-D tensor of shape[num_boxes, 4].DrawBoundingBoxes.Inputs.boxes3-D with shape[batch, num_bounding_boxes, 4]containing bounding boxes.NonMaxSuppression.Inputs.boxesA 2-D float tensor of shape[num_boxes, 4].CropAndResize.Inputs.boxIndA 1-D tensor of shape[num_boxes]with int32 values in[0, batch).CropAndResizeGradBoxes.Inputs.boxIndA 1-D tensor of shape[num_boxes]with int32 values in[0, batch).CropAndResizeGradImage.Inputs.boxIndA 1-D tensor of shape[num_boxes]with int32 values in[0, batch).NearestNeighbors.Inputs.centersMatrix of shape (m, d).DrawBoundingBoxes.Inputs.colors2-D.DecodeAndCropJpeg.Inputs.contents0-D.DecodeBmp.Inputs.contents0-D.DecodeGif.Inputs.contents0-D.DecodeImage.Inputs.contents0-D.DecodeJpeg.Inputs.contents0-D.DecodePng.Inputs.contents0-D.ExtractJpegShape.Inputs.contents0-D.AdjustContrast.Inputs.contrastFactorA float multiplier for adjusting contrast.CropAndResize.Inputs.cropSizeA 1-D tensor of 2 elements,size = [crop_height, crop_width].DecodeAndCropJpeg.Inputs.cropWindow1-D.AdjustHue.Inputs.deltaA float delta to add to the hue.ImageProjectiveTransformV3.Inputs.fillValuefloat, the value to be filled when fill_mode is constant".CropAndResizeGradBoxes.Inputs.gradsA 4-D tensor of shape[num_boxes, crop_height, crop_width, depth].CropAndResizeGradImage.Inputs.gradsA 4-D tensor of shape[num_boxes, crop_height, crop_width, depth].ResizeBicubicGrad.Inputs.grads4-D with shape[batch, height, width, channels].ResizeBilinearGrad.Inputs.grads4-D with shape[batch, height, width, channels].ResizeNearestNeighborGrad.Inputs.grads4-D with shape[batch, height, width, channels].ScaleAndTranslateGrad.Inputs.gradsThe grads inputCropAndResize.Inputs.imageA 4-D tensor of shape[batch, image_height, image_width, depth].CropAndResizeGradBoxes.Inputs.imageA 4-D tensor of shape[batch, image_height, image_width, depth].EncodeJpeg.Inputs.image3-D with shape[height, width, channels].EncodePng.Inputs.image3-D with shape[height, width, channels].RandomCrop.Inputs.image3-D of shape[height, width, channels].GenerateBoundingBoxProposals.Inputs.imageInfoA 2-D float tensor of shape[num_images, 5]containing image information Height, Width, Scale.AdjustContrast.Inputs.imagesImages to adjust.AdjustHue.Inputs.imagesImages to adjust.AdjustSaturation.Inputs.imagesImages to adjust.DrawBoundingBoxes.Inputs.images4-D with shape[batch, height, width, depth].EncodeJpegVariableQuality.Inputs.imagesImages to adjust.ExtractImagePatches.Inputs.images4-D Tensor with shape[batch, in_rows, in_cols, depth].HsvToRgb.Inputs.images1-D or higher rank.ImageProjectiveTransformV2.Inputs.images4-D with shape[batch, height, width, channels].ImageProjectiveTransformV3.Inputs.images4-D with shape[batch, height, width, channels].QuantizedResizeBilinear.Inputs.images4-D with shape[batch, height, width, channels].ResizeArea.Inputs.images4-D with shape[batch, height, width, channels].ResizeBicubic.Inputs.images4-D with shape[batch, height, width, channels].ResizeBilinear.Inputs.images4-D with shape[batch, height, width, channels].ResizeNearestNeighbor.Inputs.images4-D with shape[batch, height, width, channels].RgbToHsv.Inputs.images1-D or higher rank.ScaleAndTranslate.Inputs.imagesThe images inputCropAndResizeGradImage.Inputs.imageSizeA 1-D tensor with value[batch, image_height, image_width, depth]containing the original image size.SampleDistortedBoundingBox.Inputs.imageSize1-D, containing[height, width, channels].StatelessSampleDistortedBoundingBox.Inputs.imageSize1-D, containing[height, width, channels].ExtractGlimpse.Inputs.inputA 4-D float tensor of shape[batch_size, height, width, channels].CombinedNonMaxSuppression.Inputs.iouThresholdA 0-D float tensor representing the threshold for deciding whether boxes overlap too much with respect to IOU.NonMaxSuppression.Inputs.iouThresholdA 0-D float tensor representing the threshold for deciding whether boxes overlap too much with respect to IOU.NearestNeighbors.Inputs.kNumber of nearest centers to return for each point.QuantizedResizeBilinear.Inputs.maxThe max inputNonMaxSuppression.Inputs.maxOutputSizeA scalar integer tensor representing the maximum number of boxes to be selected by non max suppression.NonMaxSuppressionWithOverlaps.Inputs.maxOutputSizeA scalar integer tensor representing the maximum number of boxes to be selected by non max suppression.CombinedNonMaxSuppression.Inputs.maxOutputSizePerClassA scalar integer tensor representing the maximum number of boxes to be selected by non max suppression per classCombinedNonMaxSuppression.Inputs.maxTotalSizeAn int32 scalar representing the maximum number of boxes retained over all classes.QuantizedResizeBilinear.Inputs.minThe min inputSampleDistortedBoundingBox.Inputs.minObjectCoveredThe cropped area of the image must contain at least this fraction of any bounding box supplied.StatelessSampleDistortedBoundingBox.Inputs.minObjectCoveredThe cropped area of the image must contain at least this fraction of any bounding box supplied.GenerateBoundingBoxProposals.Inputs.minSizeA scalar float tensor.GenerateBoundingBoxProposals.Inputs.nmsThresholdA scalar float tensor for non-maximal-suppression threshold.ExtractGlimpse.Inputs.offsetsA 2-D integer tensor of shape[batch_size, 2]containing the y, x locations of the center of each window.ResizeBicubicGrad.Inputs.originalImage4-D with shape[batch, orig_height, orig_width, channels], The image tensor that was resized.ResizeBilinearGrad.Inputs.originalImage4-D with shape[batch, orig_height, orig_width, channels], The image tensor that was resized.ScaleAndTranslateGrad.Inputs.originalImageThe originalImage inputImageProjectiveTransformV2.Inputs.outputShape1-D Tensor [new_height, new_width].ImageProjectiveTransformV3.Inputs.outputShape1-D Tensor [new_height, new_width].NonMaxSuppressionWithOverlaps.Inputs.overlapsA 2-D float tensor of shape[num_boxes, num_boxes]representing the n-by-n box overlap values.NonMaxSuppressionWithOverlaps.Inputs.overlapThresholdA 0-D float tensor representing the threshold for deciding whether boxes overlap too.NearestNeighbors.Inputs.pointsMatrix of shape (n, d).GenerateBoundingBoxProposals.Inputs.preNmsTopnA scalar int tensor for the number of top scoring boxes to be used as input.EncodeJpegVariableQuality.Inputs.qualityAn int quality to encode to.AdjustSaturation.Inputs.scaleA float scale to add to the saturation.ScaleAndTranslate.Inputs.scaleThe scale inputScaleAndTranslateGrad.Inputs.scaleThe scale inputCombinedNonMaxSuppression.Inputs.scoresA 3-D float tensor of shape[batch_size, num_boxes, num_classes]representing a single score corresponding to each box (each row of boxes).GenerateBoundingBoxProposals.Inputs.scoresA 4-D float tensor of shape[num_images, height, width, num_achors]containing scores of the boxes for given anchors, can be unsorted.NonMaxSuppression.Inputs.scoresA 1-D float tensor of shape[num_boxes]representing a single score corresponding to each box (each row of boxes).NonMaxSuppressionWithOverlaps.Inputs.scoresA 1-D float tensor of shape[num_boxes]representing a single score corresponding to each box (each row of boxes).CombinedNonMaxSuppression.Inputs.scoreThresholdA 0-D float tensor representing the threshold for deciding when to remove boxes based on score.NonMaxSuppression.Inputs.scoreThresholdA 0-D float tensor representing the threshold for deciding when to remove boxes based on score.NonMaxSuppressionWithOverlaps.Inputs.scoreThresholdA 0-D float tensor representing the threshold for deciding when to remove boxes based on score.StatelessSampleDistortedBoundingBox.Inputs.seed1-D with shape[2].ExtractGlimpse.Inputs.sizeOutputA 1-D tensor of 2 elements containing the size of the glimpses to extract.QuantizedResizeBilinear.Inputs.sizeOutput= A 1-D int32 Tensor of 2 elements:new_height, new_width.RandomCrop.Inputs.sizeOutput1-D of length 2 containing:crop_height,crop_width..ResizeArea.Inputs.sizeOutput= A 1-D int32 Tensor of 2 elements:new_height, new_width.ResizeBicubic.Inputs.sizeOutput= A 1-D int32 Tensor of 2 elements:new_height, new_width.ResizeBilinear.Inputs.sizeOutput= A 1-D int32 Tensor of 2 elements:new_height, new_width.ResizeNearestNeighbor.Inputs.sizeOutput= A 1-D int32 Tensor of 2 elements:new_height, new_width.ResizeNearestNeighborGrad.Inputs.sizeOutput= A 1-D int32 Tensor of 2 elements:orig_height, orig_width.ScaleAndTranslate.Inputs.sizeOutputThe sizeOutput inputNonMaxSuppression.Inputs.softNmsSigmaA 0-D float tensor representing the sigma parameter for Soft NMS; see Bodla et al (c.f. https://arxiv.org/abs/1704.04503).ImageProjectiveTransformV2.Inputs.transforms2-D Tensor,[batch, 8]or[1, 8]matrix, where each row corresponds to a 3 x 3 projective transformation matrix, with the last entry assumed to be 1.ImageProjectiveTransformV3.Inputs.transforms2-D Tensor,[batch, 8]or[1, 8]matrix, where each row corresponds to a 3 x 3 projective transformation matrix, with the last entry assumed to be 1.ScaleAndTranslate.Inputs.translationThe translation inputScaleAndTranslateGrad.Inputs.translationThe translation inputMethods in org.tensorflow.op.image with parameters of type OperandModifier and TypeMethodDescriptionstatic <T extends TNumber>
AdjustContrast<T> Factory method to create a class wrapping a new AdjustContrastv2 operation.Factory method to create a class wrapping a new AdjustHue operation.static <T extends TNumber>
AdjustSaturation<T> Factory method to create a class wrapping a new AdjustSaturation operation.static CombinedNonMaxSuppressionCombinedNonMaxSuppression.create(Scope scope, Operand<TFloat32> boxes, Operand<TFloat32> scores, Operand<TInt32> maxOutputSizePerClass, Operand<TInt32> maxTotalSize, Operand<TFloat32> iouThreshold, Operand<TFloat32> scoreThreshold, CombinedNonMaxSuppression.Options... options) Factory method to create a class wrapping a new CombinedNonMaxSuppression operation.static CropAndResizeCropAndResize.create(Scope scope, Operand<? extends TNumber> image, Operand<TFloat32> boxes, Operand<TInt32> boxInd, Operand<TInt32> cropSize, CropAndResize.Options... options) Factory method to create a class wrapping a new CropAndResize operation.static CropAndResizeGradBoxesCropAndResizeGradBoxes.create(Scope scope, Operand<TFloat32> grads, Operand<? extends TNumber> image, Operand<TFloat32> boxes, Operand<TInt32> boxInd, CropAndResizeGradBoxes.Options... options) Factory method to create a class wrapping a new CropAndResizeGradBoxes operation.static <T extends TNumber>
CropAndResizeGradImage<T> CropAndResizeGradImage.create(Scope scope, Operand<TFloat32> grads, Operand<TFloat32> boxes, Operand<TInt32> boxInd, Operand<TInt32> imageSize, Class<T> T, CropAndResizeGradImage.Options... options) Factory method to create a class wrapping a new CropAndResizeGradImage operation.static DecodeAndCropJpegDecodeAndCropJpeg.create(Scope scope, Operand<TString> contents, Operand<TInt32> cropWindow, DecodeAndCropJpeg.Options... options) Factory method to create a class wrapping a new DecodeAndCropJpeg operation.static DecodeBmpDecodeBmp.create(Scope scope, Operand<TString> contents, DecodeBmp.Options... options) Factory method to create a class wrapping a new DecodeBmp operation.static DecodeGifFactory method to create a class wrapping a new DecodeGif operation.static <T extends TNumber>
DecodeImage<T> DecodeImage.create(Scope scope, Operand<TString> contents, Class<T> dtype, DecodeImage.Options... options) Factory method to create a class wrapping a new DecodeImage operation.static DecodeImage<TUint8> DecodeImage.create(Scope scope, Operand<TString> contents, DecodeImage.Options[] options) Factory method to create a class wrapping a new DecodeImage operation, with the default output types.static DecodeJpegDecodeJpeg.create(Scope scope, Operand<TString> contents, DecodeJpeg.Options... options) Factory method to create a class wrapping a new DecodeJpeg operation.DecodePng.create(Scope scope, Operand<TString> contents, Class<T> dtype, DecodePng.Options... options) Factory method to create a class wrapping a new DecodePng operation.DecodePng.create(Scope scope, Operand<TString> contents, DecodePng.Options[] options) Factory method to create a class wrapping a new DecodePng operation, with the default output types.static <T extends TNumber>
DrawBoundingBoxes<T> DrawBoundingBoxes.create(Scope scope, Operand<T> images, Operand<TFloat32> boxes, Operand<TFloat32> colors) Factory method to create a class wrapping a new DrawBoundingBoxesV2 operation.static EncodeJpegEncodeJpeg.create(Scope scope, Operand<TUint8> image, EncodeJpeg.Options... options) Factory method to create a class wrapping a new EncodeJpeg operation.static EncodeJpegVariableQualityFactory method to create a class wrapping a new EncodeJpegVariableQuality operation.static EncodePngEncodePng.create(Scope scope, Operand<? extends TNumber> image, EncodePng.Options... options) Factory method to create a class wrapping a new EncodePng operation.static ExtractGlimpseExtractGlimpse.create(Scope scope, Operand<TFloat32> input, Operand<TInt32> sizeOutput, Operand<TFloat32> offsets, ExtractGlimpse.Options... options) Factory method to create a class wrapping a new ExtractGlimpseV2 operation.static <T extends TType>
ExtractImagePatches<T> ExtractImagePatches.create(Scope scope, Operand<T> images, List<Long> ksizes, List<Long> strides, List<Long> rates, String padding) Factory method to create a class wrapping a new ExtractImagePatches operation.static ExtractJpegShape<TInt32> Factory method to create a class wrapping a new ExtractJpegShape operation, with the default output types.static <T extends TNumber>
ExtractJpegShape<T> Factory method to create a class wrapping a new ExtractJpegShape operation.static GenerateBoundingBoxProposalsGenerateBoundingBoxProposals.create(Scope scope, Operand<TFloat32> scores, Operand<TFloat32> bboxDeltas, Operand<TFloat32> imageInfo, Operand<TFloat32> anchors, Operand<TFloat32> nmsThreshold, Operand<TInt32> preNmsTopn, Operand<TFloat32> minSize, GenerateBoundingBoxProposals.Options... options) Factory method to create a class wrapping a new GenerateBoundingBoxProposals operation.Factory method to create a class wrapping a new HSVToRGB operation.static <T extends TNumber>
ImageProjectiveTransformV2<T> ImageProjectiveTransformV2.create(Scope scope, Operand<T> images, Operand<TFloat32> transforms, Operand<TInt32> outputShape, String interpolation, ImageProjectiveTransformV2.Options... options) Factory method to create a class wrapping a new ImageProjectiveTransformV2 operation.static <T extends TNumber>
ImageProjectiveTransformV3<T> ImageProjectiveTransformV3.create(Scope scope, Operand<T> images, Operand<TFloat32> transforms, Operand<TInt32> outputShape, Operand<TFloat32> fillValue, String interpolation, ImageProjectiveTransformV3.Options... options) Factory method to create a class wrapping a new ImageProjectiveTransformV3 operation.static NearestNeighborsNearestNeighbors.create(Scope scope, Operand<TFloat32> points, Operand<TFloat32> centers, Operand<TInt64> k) Factory method to create a class wrapping a new NearestNeighbors operation.static <T extends TNumber>
NonMaxSuppression<T> NonMaxSuppression.create(Scope scope, Operand<T> boxes, Operand<T> scores, Operand<TInt32> maxOutputSize, Operand<T> iouThreshold, Operand<T> scoreThreshold, Operand<T> softNmsSigma, NonMaxSuppression.Options... options) Factory method to create a class wrapping a new NonMaxSuppressionV5 operation.NonMaxSuppressionWithOverlaps.create(Scope scope, Operand<TFloat32> overlaps, Operand<TFloat32> scores, Operand<TInt32> maxOutputSize, Operand<TFloat32> overlapThreshold, Operand<TFloat32> scoreThreshold) Factory method to create a class wrapping a new NonMaxSuppressionWithOverlaps operation.static <T extends TNumber>
QuantizedResizeBilinear<T> QuantizedResizeBilinear.create(Scope scope, Operand<T> images, Operand<TInt32> sizeOutput, Operand<TFloat32> min, Operand<TFloat32> max, QuantizedResizeBilinear.Options... options) Factory method to create a class wrapping a new QuantizedResizeBilinear operation.static <T extends TNumber>
RandomCrop<T> RandomCrop.create(Scope scope, Operand<T> image, Operand<TInt64> sizeOutput, RandomCrop.Options... options) Factory method to create a class wrapping a new RandomCrop operation.static ResizeAreaResizeArea.create(Scope scope, Operand<? extends TNumber> images, Operand<TInt32> sizeOutput, ResizeArea.Options... options) Factory method to create a class wrapping a new ResizeArea operation.static ResizeBicubicResizeBicubic.create(Scope scope, Operand<? extends TNumber> images, Operand<TInt32> sizeOutput, ResizeBicubic.Options... options) Factory method to create a class wrapping a new ResizeBicubic operation.static <T extends TNumber>
ResizeBicubicGrad<T> ResizeBicubicGrad.create(Scope scope, Operand<TFloat32> grads, Operand<T> originalImage, ResizeBicubicGrad.Options... options) Factory method to create a class wrapping a new ResizeBicubicGrad operation.static ResizeBilinearResizeBilinear.create(Scope scope, Operand<? extends TNumber> images, Operand<TInt32> sizeOutput, ResizeBilinear.Options... options) Factory method to create a class wrapping a new ResizeBilinear operation.static <T extends TNumber>
ResizeBilinearGrad<T> ResizeBilinearGrad.create(Scope scope, Operand<TFloat32> grads, Operand<T> originalImage, ResizeBilinearGrad.Options... options) Factory method to create a class wrapping a new ResizeBilinearGrad operation.static <T extends TNumber>
ResizeNearestNeighbor<T> ResizeNearestNeighbor.create(Scope scope, Operand<T> images, Operand<TInt32> sizeOutput, ResizeNearestNeighbor.Options... options) Factory method to create a class wrapping a new ResizeNearestNeighbor operation.static <T extends TNumber>
ResizeNearestNeighborGrad<T> ResizeNearestNeighborGrad.create(Scope scope, Operand<T> grads, Operand<TInt32> sizeOutput, ResizeNearestNeighborGrad.Options... options) Factory method to create a class wrapping a new ResizeNearestNeighborGrad operation.Factory method to create a class wrapping a new RGBToHSV operation.static <T extends TNumber>
SampleDistortedBoundingBox<T> SampleDistortedBoundingBox.create(Scope scope, Operand<T> imageSize, Operand<TFloat32> boundingBoxes, Operand<TFloat32> minObjectCovered, SampleDistortedBoundingBox.Options... options) Factory method to create a class wrapping a new SampleDistortedBoundingBoxV2 operation.static ScaleAndTranslateScaleAndTranslate.create(Scope scope, Operand<? extends TNumber> images, Operand<TInt32> sizeOutput, Operand<TFloat32> scale, Operand<TFloat32> translation, ScaleAndTranslate.Options... options) Factory method to create a class wrapping a new ScaleAndTranslate operation.static <T extends TNumber>
ScaleAndTranslateGrad<T> ScaleAndTranslateGrad.create(Scope scope, Operand<T> grads, Operand<T> originalImage, Operand<TFloat32> scale, Operand<TFloat32> translation, ScaleAndTranslateGrad.Options... options) Factory method to create a class wrapping a new ScaleAndTranslateGrad operation.static <T extends TNumber>
StatelessSampleDistortedBoundingBox<T> StatelessSampleDistortedBoundingBox.create(Scope scope, Operand<T> imageSize, Operand<TFloat32> boundingBoxes, Operand<TFloat32> minObjectCovered, Operand<? extends TNumber> seed, StatelessSampleDistortedBoundingBox.Options... options) Factory method to create a class wrapping a new StatelessSampleDistortedBoundingBox operation. -
Uses of Operand in org.tensorflow.op.io
Classes in org.tensorflow.op.io that implement OperandModifier and TypeClassDescriptionfinal classDecode web-safe base64-encoded strings.final classDecompress strings.final classConvert JSON-encoded Example records to binary protocol buffer strings.final classDecodePaddedRaw<T extends TNumber>Reinterpret the bytes of a string as a vector of numbers.final classReinterpret the bytes of a string as a vector of numbers.final classEncode strings into web-safe base64 format.final classDeprecated.final classA queue that produces elements in first-in first-out order.final classA Reader that outputs fixed-length records from a file.final classA Reader that outputs the queued work as both the key and value.final classA Reader that outputs the records from a LMDB file.final classReturns the set of files matching one or more glob patterns.final classA queue that produces elements in first-in first-out order.final classParseTensor<T extends TType>Transforms a serialized tensorflow.TensorProto proto into a Tensor.final classA queue that produces elements sorted by the first component value.final classReturns true if queue is closed.final classComputes the number of elements in the given queue.final classA queue that randomizes the order of elements.final classReturns the number of records this Reader has produced.final classReturns the number of work units this Reader has finished processing.final classProduce a string tensor that encodes the state of a Reader.final classReads and outputs the entire contents of the input filename.final classSerializeManySparse<U extends TType>Serialize anN-minibatchSparseTensorinto an[N, 3]Tensorobject.final classSerializeSparse<U extends TType>Serialize aSparseTensorinto a[3]Tensorobject.final classTransforms a Tensor into a serialized TensorProto proto.final classGenerate a sharded filename.final classGenerate a glob pattern matching all sharded file names.final classA Reader that outputs the lines of a file delimited by '\n'.final classA Reader that outputs the records from a TensorFlow Records file.final classA Reader that outputs the entire contents of a file as a value.Classes in org.tensorflow.op.io that implement interfaces with type arguments of type OperandModifier and TypeClassDescriptionfinal classConvert CSV records to tensors.final classDequeues a tuple of one or more tensors from the given queue.final classDequeuesntuples of one or more tensors from the given queue.final classDequeuesntuples of one or more tensors from the given queue.Fields in org.tensorflow.op.io declared as OperandModifier and TypeFieldDescriptionShardedFilename.Inputs.basenameThe basename inputShardedFilespec.Inputs.basenameThe basename inputDecodeCompressed.Inputs.bytesA Tensor of string which is compressed.DecodeRaw.Inputs.bytesAll the elements must have the same length.WriteFile.Inputs.contentsscalar.ParseSequenceExample.Inputs.contextDenseKeysThe keys expected in the SequenceExamples' context features associated with dense values.ParseSequenceExample.Inputs.contextRaggedKeysThe keys expected in the Examples' features associated with context_ragged values.ParseSequenceExample.Inputs.contextSparseKeysThe keys expected in the Examples' features associated with context_sparse values.ParseSequenceExample.Inputs.debugNameA scalar or vector containing the names of the serialized protos.ParseSingleSequenceExample.Inputs.debugNameA scalar containing the name of the serialized proto.ParseExample.Inputs.denseKeysVector of strings.ParseSequenceExample.Inputs.featureListDenseKeysThe keys expected in the SequenceExamples' feature_lists associated with lists of dense values.ParseSequenceExample.Inputs.featureListDenseMissingAssumedEmptyA vector corresponding 1:1 with feature_list_dense_keys, indicating which features may be missing from the SequenceExamples.ParseSingleSequenceExample.Inputs.featureListDenseMissingAssumedEmptyA vector listing the FeatureList keys which may be missing from the SequenceExample.ParseSequenceExample.Inputs.featureListRaggedKeysThe keys expected in the FeatureLists associated with ragged values.ParseSequenceExample.Inputs.featureListSparseKeysThe keys expected in the FeatureLists associated with sparse values.ReadFile.Inputs.filenameThe filename inputWriteFile.Inputs.filenamescalar.DecodePaddedRaw.Inputs.fixedLengthLength in bytes for each element of the decoded output.QueueClose.Inputs.handleThe handle to a queue.QueueDequeue.Inputs.handleThe handle to a queue.QueueDequeueMany.Inputs.handleThe handle to a queue.QueueDequeueUpTo.Inputs.handleThe handle to a queue.QueueEnqueue.Inputs.handleThe handle to a queue.QueueEnqueueMany.Inputs.handleThe handle to a queue.QueueIsClosed.Inputs.handleThe handle to a queue.QueueSize.Inputs.handleThe handle to a queue.DecodeBase64.Inputs.inputBase64 strings to decode.EncodeBase64.Inputs.inputStrings to be encoded.DecodePaddedRaw.Inputs.inputBytesTensor of string to be decoded.DecodeJsonExample.Inputs.jsonExamplesEach string is a JSON object serialized according to the JSON mapping of the Example proto.QueueDequeueMany.Inputs.nThe number of tuples to dequeue.QueueDequeueUpTo.Inputs.nThe number of tuples to dequeue.ParseExample.Inputs.namesA tensor containing the names of the serialized protos.ReaderReadUpTo.Inputs.numRecordsnumber of records to read fromReader.ShardedFilename.Inputs.numShardsThe numShards inputShardedFilespec.Inputs.numShardsThe numShards inputMatchingFiles.Inputs.patternShell wildcard pattern(s).ReaderRead.Inputs.queueHandleHandle to a Queue, with string work items.ReaderReadUpTo.Inputs.queueHandleHandle to aQueue, with string work items.ParseExample.Inputs.raggedKeysVector of strings.ReaderNumRecordsProduced.Inputs.readerHandleHandle to a Reader.ReaderNumWorkUnitsCompleted.Inputs.readerHandleHandle to a Reader.ReaderRead.Inputs.readerHandleHandle to a Reader.ReaderReadUpTo.Inputs.readerHandleHandle to aReader.ReaderReset.Inputs.readerHandleHandle to a Reader.ReaderRestoreState.Inputs.readerHandleHandle to a Reader.ReaderSerializeState.Inputs.readerHandleHandle to a Reader.DecodeCsv.Inputs.recordsEach string is a record/row in the csv and all records should have the same format.DisableCopyOnRead.Inputs.resourceThe resource handle of the resource variable.FakeQueue.Inputs.resourceThe resource inputParseExample.Inputs.serializedA scalar or vector containing binary serialized Example protos.ParseSequenceExample.Inputs.serializedA scalar or vector containing binary serialized SequenceExample protos.ParseSingleExample.Inputs.serializedA vector containing a batch of binary serialized Example protos.ParseSingleSequenceExample.Inputs.serializedA scalar containing a binary serialized SequenceExample proto.ParseTensor.Inputs.serializedA scalar string containing a serialized TensorProto proto.DeserializeManySparse.Inputs.serializedSparse2-D, TheNserializedSparseTensorobjects.ShardedFilename.Inputs.shardThe shard inputSerializeManySparse.Inputs.sparseIndices2-D.SerializeSparse.Inputs.sparseIndices2-D.ParseExample.Inputs.sparseKeysVector of strings.SerializeManySparse.Inputs.sparseShape1-D.SerializeSparse.Inputs.sparseShape1-D.SerializeManySparse.Inputs.sparseValues1-D.SerializeSparse.Inputs.sparseValues1-D.ReaderRestoreState.Inputs.stateResult of a ReaderSerializeState of a Reader with type matching reader_handle.SerializeTensor.Inputs.tensorA Tensor of typeT.Fields in org.tensorflow.op.io with type parameters of type OperandModifier and TypeFieldDescriptionQueueEnqueue.Inputs.componentsOne or more tensors from which the enqueued tensors should be taken.QueueEnqueueMany.Inputs.componentsOne or more tensors from which the enqueued tensors should be taken.ParseSequenceExample.Inputs.contextDenseDefaultsA list of Ncontext_dense Tensors (some may be empty). context_dense_defaults[j] provides default values when the SequenceExample's context map lacks context_dense_key[j].ParseSingleSequenceExample.Inputs.contextDenseDefaultsA list of Ncontext_dense Tensors (some may be empty). context_dense_defaults[j] provides default values when the SequenceExample's context map lacks context_dense_key[j].ParseSingleSequenceExample.Inputs.contextDenseKeysA list of Ncontext_dense string Tensors (scalars).ParseSingleSequenceExample.Inputs.contextSparseKeysA list of Ncontext_sparse string Tensors (scalars).ParseExample.Inputs.denseDefaultsA list of Tensors (some may be empty).ParseSingleExample.Inputs.denseDefaultsA list of Tensors (some may be empty), whose length matches the length ofdense_keys. dense_defaults[j] provides default values when the example's feature_map lacks dense_key[j].ParseSingleSequenceExample.Inputs.featureListDenseKeysA list of Nfeature_list_dense string Tensors (scalars).ParseSingleSequenceExample.Inputs.featureListSparseKeysA list of Nfeature_list_sparse string Tensors (scalars).DecodeCsv.Inputs.recordDefaultsOne tensor per column of the input record, with either a scalar default value for that column or an empty vector if the column is required.Methods in org.tensorflow.op.io that return types with arguments of type OperandModifier and TypeMethodDescriptionDecodeCsv.iterator()QueueDequeue.iterator()QueueDequeueMany.iterator()QueueDequeueUpTo.iterator()Methods in org.tensorflow.op.io with parameters of type OperandModifier and TypeMethodDescriptionstatic DecodeBase64Factory method to create a class wrapping a new DecodeBase64 operation.static DecodeCompressedDecodeCompressed.create(Scope scope, Operand<TString> bytes, DecodeCompressed.Options... options) Factory method to create a class wrapping a new DecodeCompressed operation.static DecodeCsvDecodeCsv.create(Scope scope, Operand<TString> records, Iterable<Operand<?>> recordDefaults, DecodeCsv.Options... options) Factory method to create a class wrapping a new DecodeCSV operation.static DecodeJsonExampleFactory method to create a class wrapping a new DecodeJSONExample operation.static <T extends TNumber>
DecodePaddedRaw<T> DecodePaddedRaw.create(Scope scope, Operand<TString> inputBytes, Operand<TInt32> fixedLength, Class<T> outType, DecodePaddedRaw.Options... options) Factory method to create a class wrapping a new DecodePaddedRaw operation.DecodeRaw.create(Scope scope, Operand<TString> bytes, Class<T> outType, DecodeRaw.Options... options) Factory method to create a class wrapping a new DecodeRaw operation.static <T extends TType>
DeserializeManySparse<T> Factory method to create a class wrapping a new DeserializeManySparse operation.static DisableCopyOnReadFactory method to create a class wrapping a new DisableCopyOnRead operation.static EncodeBase64EncodeBase64.create(Scope scope, Operand<TString> input, EncodeBase64.Options... options) Factory method to create a class wrapping a new EncodeBase64 operation.static FakeQueueFactory method to create a class wrapping a new FakeQueue operation.static MatchingFilesFactory method to create a class wrapping a new MatchingFiles operation.static ParseExampleParseExample.create(Scope scope, Operand<TString> serialized, Operand<TString> names, Operand<TString> sparseKeys, Operand<TString> denseKeys, Operand<TString> raggedKeys, Iterable<Operand<?>> denseDefaults, Long numSparse, List<Class<? extends TType>> sparseTypes, List<Class<? extends TType>> raggedValueTypes, List<Class<? extends TNumber>> raggedSplitTypes, List<Shape> denseShapes) Factory method to create a class wrapping a new ParseExampleV2 operation.static ParseSequenceExampleParseSequenceExample.create(Scope scope, Operand<TString> serialized, Operand<TString> debugName, Operand<TString> contextSparseKeys, Operand<TString> contextDenseKeys, Operand<TString> contextRaggedKeys, Operand<TString> featureListSparseKeys, Operand<TString> featureListDenseKeys, Operand<TString> featureListRaggedKeys, Operand<TBool> featureListDenseMissingAssumedEmpty, Iterable<Operand<?>> contextDenseDefaults, List<Class<? extends TType>> contextSparseTypes, List<Class<? extends TType>> contextRaggedValueTypes, List<Class<? extends TNumber>> contextRaggedSplitTypes, List<Class<? extends TType>> featureListDenseTypes, List<Class<? extends TType>> featureListSparseTypes, List<Class<? extends TType>> featureListRaggedValueTypes, List<Class<? extends TNumber>> featureListRaggedSplitTypes, ParseSequenceExample.Options... options) Factory method to create a class wrapping a new ParseSequenceExampleV2 operation.static ParseSingleExampleParseSingleExample.create(Scope scope, Operand<TString> serialized, Iterable<Operand<?>> denseDefaults, Long numSparse, List<String> sparseKeys, List<String> denseKeys, List<Class<? extends TType>> sparseTypes, List<Shape> denseShapes) Factory method to create a class wrapping a new ParseSingleExample operation.static ParseSingleSequenceExampleParseSingleSequenceExample.create(Scope scope, Operand<TString> serialized, Operand<TString> featureListDenseMissingAssumedEmpty, Iterable<Operand<TString>> contextSparseKeys, Iterable<Operand<TString>> contextDenseKeys, Iterable<Operand<TString>> featureListSparseKeys, Iterable<Operand<TString>> featureListDenseKeys, Iterable<Operand<?>> contextDenseDefaults, Operand<TString> debugName, List<Class<? extends TType>> contextSparseTypes, List<Class<? extends TType>> featureListDenseTypes, List<Class<? extends TType>> featureListSparseTypes, ParseSingleSequenceExample.Options... options) Factory method to create a class wrapping a new ParseSingleSequenceExample operation.static <T extends TType>
ParseTensor<T> Factory method to create a class wrapping a new ParseTensor operation.static QueueCloseQueueClose.create(Scope scope, Operand<? extends TType> handle, QueueClose.Options... options) Factory method to create a class wrapping a new QueueCloseV2 operation.static QueueDequeueQueueDequeue.create(Scope scope, Operand<? extends TType> handle, List<Class<? extends TType>> componentTypes, QueueDequeue.Options... options) Factory method to create a class wrapping a new QueueDequeueV2 operation.static QueueDequeueManyQueueDequeueMany.create(Scope scope, Operand<? extends TType> handle, Operand<TInt32> n, List<Class<? extends TType>> componentTypes, QueueDequeueMany.Options... options) Factory method to create a class wrapping a new QueueDequeueManyV2 operation.static QueueDequeueUpToQueueDequeueUpTo.create(Scope scope, Operand<? extends TType> handle, Operand<TInt32> n, List<Class<? extends TType>> componentTypes, QueueDequeueUpTo.Options... options) Factory method to create a class wrapping a new QueueDequeueUpToV2 operation.static QueueEnqueueQueueEnqueue.create(Scope scope, Operand<? extends TType> handle, Iterable<Operand<?>> components, QueueEnqueue.Options... options) Factory method to create a class wrapping a new QueueEnqueueV2 operation.static QueueEnqueueManyQueueEnqueueMany.create(Scope scope, Operand<? extends TType> handle, Iterable<Operand<?>> components, QueueEnqueueMany.Options... options) Factory method to create a class wrapping a new QueueEnqueueManyV2 operation.static QueueIsClosedFactory method to create a class wrapping a new QueueIsClosedV2 operation.static QueueSizeFactory method to create a class wrapping a new QueueSizeV2 operation.static ReaderNumRecordsProducedFactory method to create a class wrapping a new ReaderNumRecordsProducedV2 operation.static ReaderNumWorkUnitsCompletedFactory method to create a class wrapping a new ReaderNumWorkUnitsCompletedV2 operation.static ReaderReadReaderRead.create(Scope scope, Operand<? extends TType> readerHandle, Operand<? extends TType> queueHandle) Factory method to create a class wrapping a new ReaderReadV2 operation.static ReaderReadUpToReaderReadUpTo.create(Scope scope, Operand<? extends TType> readerHandle, Operand<? extends TType> queueHandle, Operand<TInt64> numRecords) Factory method to create a class wrapping a new ReaderReadUpToV2 operation.static ReaderResetFactory method to create a class wrapping a new ReaderResetV2 operation.static ReaderRestoreStateReaderRestoreState.create(Scope scope, Operand<? extends TType> readerHandle, Operand<TString> state) Factory method to create a class wrapping a new ReaderRestoreStateV2 operation.static ReaderSerializeStateFactory method to create a class wrapping a new ReaderSerializeStateV2 operation.static ReadFileFactory method to create a class wrapping a new ReadFile operation.static SerializeManySparse<TString> SerializeManySparse.create(Scope scope, Operand<TInt64> sparseIndices, Operand<? extends TType> sparseValues, Operand<TInt64> sparseShape) Factory method to create a class wrapping a new SerializeManySparse operation, with the default output types.static <U extends TType>
SerializeManySparse<U> SerializeManySparse.create(Scope scope, Operand<TInt64> sparseIndices, Operand<? extends TType> sparseValues, Operand<TInt64> sparseShape, Class<U> outType) Factory method to create a class wrapping a new SerializeManySparse operation.static SerializeSparse<TString> SerializeSparse.create(Scope scope, Operand<TInt64> sparseIndices, Operand<? extends TType> sparseValues, Operand<TInt64> sparseShape) Factory method to create a class wrapping a new SerializeSparse operation, with the default output types.static <U extends TType>
SerializeSparse<U> SerializeSparse.create(Scope scope, Operand<TInt64> sparseIndices, Operand<? extends TType> sparseValues, Operand<TInt64> sparseShape, Class<U> outType) Factory method to create a class wrapping a new SerializeSparse operation.static SerializeTensorFactory method to create a class wrapping a new SerializeTensor operation.static ShardedFilenameShardedFilename.create(Scope scope, Operand<TString> basename, Operand<TInt32> shard, Operand<TInt32> numShards) Factory method to create a class wrapping a new ShardedFilename operation.static ShardedFilespecFactory method to create a class wrapping a new ShardedFilespec operation.static WriteFileFactory method to create a class wrapping a new WriteFile operation.Method parameters in org.tensorflow.op.io with type arguments of type OperandModifier and TypeMethodDescriptionstatic DecodeCsvDecodeCsv.create(Scope scope, Operand<TString> records, Iterable<Operand<?>> recordDefaults, DecodeCsv.Options... options) Factory method to create a class wrapping a new DecodeCSV operation.static ParseExampleParseExample.create(Scope scope, Operand<TString> serialized, Operand<TString> names, Operand<TString> sparseKeys, Operand<TString> denseKeys, Operand<TString> raggedKeys, Iterable<Operand<?>> denseDefaults, Long numSparse, List<Class<? extends TType>> sparseTypes, List<Class<? extends TType>> raggedValueTypes, List<Class<? extends TNumber>> raggedSplitTypes, List<Shape> denseShapes) Factory method to create a class wrapping a new ParseExampleV2 operation.static ParseSequenceExampleParseSequenceExample.create(Scope scope, Operand<TString> serialized, Operand<TString> debugName, Operand<TString> contextSparseKeys, Operand<TString> contextDenseKeys, Operand<TString> contextRaggedKeys, Operand<TString> featureListSparseKeys, Operand<TString> featureListDenseKeys, Operand<TString> featureListRaggedKeys, Operand<TBool> featureListDenseMissingAssumedEmpty, Iterable<Operand<?>> contextDenseDefaults, List<Class<? extends TType>> contextSparseTypes, List<Class<? extends TType>> contextRaggedValueTypes, List<Class<? extends TNumber>> contextRaggedSplitTypes, List<Class<? extends TType>> featureListDenseTypes, List<Class<? extends TType>> featureListSparseTypes, List<Class<? extends TType>> featureListRaggedValueTypes, List<Class<? extends TNumber>> featureListRaggedSplitTypes, ParseSequenceExample.Options... options) Factory method to create a class wrapping a new ParseSequenceExampleV2 operation.static ParseSingleExampleParseSingleExample.create(Scope scope, Operand<TString> serialized, Iterable<Operand<?>> denseDefaults, Long numSparse, List<String> sparseKeys, List<String> denseKeys, List<Class<? extends TType>> sparseTypes, List<Shape> denseShapes) Factory method to create a class wrapping a new ParseSingleExample operation.static ParseSingleSequenceExampleParseSingleSequenceExample.create(Scope scope, Operand<TString> serialized, Operand<TString> featureListDenseMissingAssumedEmpty, Iterable<Operand<TString>> contextSparseKeys, Iterable<Operand<TString>> contextDenseKeys, Iterable<Operand<TString>> featureListSparseKeys, Iterable<Operand<TString>> featureListDenseKeys, Iterable<Operand<?>> contextDenseDefaults, Operand<TString> debugName, List<Class<? extends TType>> contextSparseTypes, List<Class<? extends TType>> featureListDenseTypes, List<Class<? extends TType>> featureListSparseTypes, ParseSingleSequenceExample.Options... options) Factory method to create a class wrapping a new ParseSingleSequenceExample operation.static QueueEnqueueQueueEnqueue.create(Scope scope, Operand<? extends TType> handle, Iterable<Operand<?>> components, QueueEnqueue.Options... options) Factory method to create a class wrapping a new QueueEnqueueV2 operation.static QueueEnqueueManyQueueEnqueueMany.create(Scope scope, Operand<? extends TType> handle, Iterable<Operand<?>> components, QueueEnqueueMany.Options... options) Factory method to create a class wrapping a new QueueEnqueueManyV2 operation. -
Uses of Operand in org.tensorflow.op.linalg
Classes in org.tensorflow.op.linalg that implement OperandModifier and TypeClassDescriptionfinal classBandedTriangularSolve<T extends TType>The BandedTriangularSolve operationfinal classCopy a tensor setting everything outside a central band in each innermost matrix to zero.final classBatchCholesky<T extends TNumber>The BatchCholesky operationfinal classBatchCholeskyGrad<T extends TNumber>The BatchCholeskyGrad operationfinal classBatchMatrixBandPart<T extends TType>The BatchMatrixBandPart operationfinal classBatchMatrixDeterminant<T extends TType>The BatchMatrixDeterminant operationfinal classBatchMatrixDiag<T extends TType>The BatchMatrixDiag operationfinal classBatchMatrixDiagPart<T extends TType>The BatchMatrixDiagPart operationfinal classBatchMatrixInverse<T extends TNumber>The BatchMatrixInverse operation DEPRECATED: This operation is deprecated and will be removed in a future version.final classBatchMatrixSetDiag<T extends TType>The BatchMatrixSetDiag operationfinal classBatchMatrixSolve<T extends TNumber>The BatchMatrixSolve operationfinal classBatchMatrixSolveLs<T extends TNumber>The BatchMatrixSolveLs operationfinal classBatchMatrixTriangularSolve<T extends TNumber>The BatchMatrixTriangularSolve operationfinal classComputes the Cholesky decomposition of one or more square matrices.final classCholeskyGrad<T extends TNumber>Computes the reverse mode backpropagated gradient of the Cholesky algorithm.final classConjugateTranspose<T extends TType>Shuffle dimensions of x according to a permutation and conjugate the result.final classCompute the pairwise cross product.final classComputes the determinant of one or more square matrices.final classTensor contraction according to Einstein summation convention.final classEuclideanNorm<T extends TType>Computes the euclidean norm of elements across dimensions of a tensor.final classComputes the inverse of one or more square invertible matrices or their adjoints (conjugate transposes).final classLoads a 2-D (matrix)Tensorwith nameold_tensor_namefrom the checkpoint atckpt_pathand potentially reorders its rows and columns using the specified remappings.final classMultiply the matrix "a" by the matrix "b".final classMatrixDiag<T extends TType>Returns a batched diagonal tensor with given batched diagonal values.final classMatrixDiagPart<T extends TType>Returns the batched diagonal part of a batched tensor.final classMatrixDiagPartV3<T extends TType>Returns the batched diagonal part of a batched tensor.final classMatrixDiagV3<T extends TType>Returns a batched diagonal tensor with given batched diagonal values.final classMatrixExponential<T extends TType>Deprecated, use python implementation tf.linalg.matrix_exponential.final classMatrixLogarithm<T extends TType>Computes the matrix logarithm of one or more square matrices: \(log(exp(A)) = A\)final classMatrixSetDiag<T extends TType>Returns a batched matrix tensor with new batched diagonal values.final classMatrixSolveLs<T extends TType>Solves one or more linear least-squares problems.final classSolves systems of linear equations.final classComputes the matrix square root of one or more square matrices: matmul(sqrtm(A), sqrtm(A)) = Afinal classTensorDiag<T extends TType>Returns a diagonal tensor with a given diagonal values.final classTensorDiagPart<T extends TType>Returns the diagonal part of the tensor.final classShuffle dimensions of x according to a permutation.final classTriangularSolve<T extends TType>Solves systems of linear equations with upper or lower triangular matrices by backsubstitution.final classTridiagonalMatMul<T extends TType>Calculate product with tridiagonal matrix.final classTridiagonalSolve<T extends TType>Solves tridiagonal systems of equations.Fields in org.tensorflow.op.linalg declared as OperandModifier and TypeFieldDescriptionCross.Inputs.aA tensor containing 3-element vectors.MatMul.Inputs.aThe a inputQuantizedMatMul.Inputs.aMust be a two-dimensional tensor.QuantizedMatMulWithBias.Inputs.aA matrix to be multiplied.QuantizedMatMulWithBiasAndRelu.Inputs.aA matrix to be multiplied.QuantizedMatMulWithBiasAndReluAndRequantize.Inputs.aA matrix to be multiplied.EuclideanNorm.Inputs.axisThe dimensions to reduce.Cross.Inputs.bAnother tensor, of same type and shape asa.MatMul.Inputs.bThe b inputQuantizedMatMul.Inputs.bMust be a two-dimensional tensor.QuantizedMatMulWithBias.Inputs.bA matrix to be multiplied and must be a two-dimensional tensor of typeqint8.QuantizedMatMulWithBiasAndRelu.Inputs.bA matrix to be multiplied and must be a two-dimensional tensor of typeqint8.QuantizedMatMulWithBiasAndReluAndRequantize.Inputs.bA matrix to be multiplied and must be a two-dimensional tensor of typeqint8.QuantizedMatMulWithBias.Inputs.biasA 1D bias tensor with size matching inner dimension ofb(after being transposed iftransposed_bis non-zero).QuantizedMatMulWithBiasAndRelu.Inputs.biasA 1D bias tensor with size matching with inner dimension ofb(after being transposed iftransposed_bis non-zero).QuantizedMatMulWithBiasAndReluAndRequantize.Inputs.biasA 1D bias tensor with size matching with inner dimension ofb(after being transposed iftransposed_bis non-zero).LoadAndRemapMatrix.Inputs.ckptPathPath to the TensorFlow checkpoint (version 2,TensorBundle) from which the old matrixTensorwill be loaded.LoadAndRemapMatrix.Inputs.colRemappingAn intTensorof column remappings (generally created bygenerate_vocab_remapping).BatchMatrixDiag.Inputs.diagonalThe diagonal inputBatchMatrixSetDiag.Inputs.diagonalThe diagonal inputMatrixDiag.Inputs.diagonalRankr, wherer >= 1MatrixDiagV3.Inputs.diagonalRankr, wherer >= 1MatrixSetDiag.Inputs.diagonalRankrwhenkis an integer ork[0] == k[1].TensorDiag.Inputs.diagonalRank k tensor where k is at most 1.TridiagonalSolve.Inputs.diagonalsTensor of shape[..., 3, M]whose innermost 2 dimensions represent the tridiagonal matrices with three rows being the superdiagonal, diagonals, and subdiagonals, in order.BatchCholeskyGrad.Inputs.gradThe grad inputCholeskyGrad.Inputs.graddf/dl where f is some scalar function.LoadAndRemapMatrix.Inputs.initializingValuesA floatTensorcontaining values to fill in for cells in the output matrix that are not loaded from the checkpoint.BandPart.Inputs.inputRankktensor.BatchCholesky.Inputs.inputThe input inputBatchMatrixBandPart.Inputs.inputThe input inputBatchMatrixDeterminant.Inputs.inputThe input inputBatchMatrixDiagPart.Inputs.inputThe input inputBatchMatrixInverse.Inputs.inputThe input inputBatchMatrixSetDiag.Inputs.inputThe input inputBatchSelfAdjointEig.Inputs.inputThe input inputBatchSvd.Inputs.inputThe input inputCholesky.Inputs.inputShape is[..., M, M].Det.Inputs.inputShape is[..., M, M].Eig.Inputs.inputTensorinput of shape[N, N].EuclideanNorm.Inputs.inputThe tensor to reduce.Inv.Inputs.inputShape is[..., M, M].LogMatrixDeterminant.Inputs.inputShape is[N, M, M].Lu.Inputs.inputA tensor of shape[..., M, M]whose inner-most 2 dimensions form matrices of size[M, M].MatrixDiagPart.Inputs.inputRankrtensor wherer >= 2.MatrixDiagPartV3.Inputs.inputRankrtensor wherer >= 2.MatrixExponential.Inputs.inputThe input inputMatrixLogarithm.Inputs.inputShape is[..., M, M].MatrixSetDiag.Inputs.inputRankr+1, wherer >= 1.Qr.Inputs.inputA tensor of shape[..., M, N]whose inner-most 2 dimensions form matrices of size[M, N].SelfAdjointEig.Inputs.inputTensorinput of shape[N, N].Sqrtm.Inputs.inputShape is[..., M, M].Svd.Inputs.inputA tensor of shape[..., M, N]whose inner-most 2 dimensions form matrices of size[M, N].TensorDiagPart.Inputs.inputRank k tensor where k is even and not zero.MatrixDiag.Inputs.kDiagonal offset(s).MatrixDiagPart.Inputs.kDiagonal offset(s).MatrixDiagPartV3.Inputs.kDiagonal offset(s).MatrixDiagV3.Inputs.kDiagonal offset(s).MatrixSetDiag.Inputs.kDiagonal offset(s).BatchCholeskyGrad.Inputs.lThe l inputCholeskyGrad.Inputs.lOutput of batch Cholesky algorithm l = cholesky(A).BatchMatrixSolveLs.Inputs.l2RegularizerThe l2Regularizer inputMatrixSolveLs.Inputs.l2RegularizerScalar tensor.TridiagonalMatMul.Inputs.maindiagTensor of shape[..., 1, M], representing main diagonals of tri-diagonal matrices to the left of multiplication.BandedTriangularSolve.Inputs.matrixThe matrix inputBatchMatrixSolve.Inputs.matrixThe matrix inputBatchMatrixSolveLs.Inputs.matrixThe matrix inputBatchMatrixTriangularSolve.Inputs.matrixThe matrix inputMatrixSolveLs.Inputs.matrixShape is[..., M, N].Solve.Inputs.matrixShape is[..., M, M].TriangularSolve.Inputs.matrixShape is[..., M, M].QuantizedMatMul.Inputs.maxAThe float value that the highest quantizedavalue represents.QuantizedMatMulWithBias.Inputs.maxAThe float value that the highest quantizedavalue represents.QuantizedMatMulWithBiasAndRelu.Inputs.maxAThe float value that the highest quantizedavalue represents.QuantizedMatMulWithBiasAndReluAndRequantize.Inputs.maxAThe float value that the highest quantizedavalue represents.QuantizedMatMul.Inputs.maxBThe float value that the highest quantizedbvalue represents.QuantizedMatMulWithBias.Inputs.maxBThe float value that the highest quantizedbvalue represents.QuantizedMatMulWithBiasAndRelu.Inputs.maxBThe float value that the highest quantizedbvalue represents.QuantizedMatMulWithBiasAndReluAndRequantize.Inputs.maxBThe float value that the highest quantizedbvalue represents.QuantizedMatMulWithBiasAndReluAndRequantize.Inputs.maxFreezedOutputThe maxFreezedOutput inputQuantizedMatMul.Inputs.minAThe float value that the lowest quantizedavalue represents.QuantizedMatMulWithBias.Inputs.minAThe float value that the lowest quantizedavalue represents.QuantizedMatMulWithBiasAndRelu.Inputs.minAThe float value that the lowest quantizedavalue represents.QuantizedMatMulWithBiasAndReluAndRequantize.Inputs.minAThe float value that the lowest quantizedavalue represents.QuantizedMatMul.Inputs.minBThe float value that the lowest quantizedbvalue represents.QuantizedMatMulWithBias.Inputs.minBThe float value that the lowest quantizedbvalue represents.QuantizedMatMulWithBiasAndRelu.Inputs.minBThe float value that the lowest quantizedbvalue represents.QuantizedMatMulWithBiasAndReluAndRequantize.Inputs.minBThe float value that the lowest quantizedbvalue represents.QuantizedMatMulWithBiasAndReluAndRequantize.Inputs.minFreezedOutputThe float value that the highest quantized output value after requantize.MatrixDiag.Inputs.numColsThe number of columns of the output matrix.MatrixDiagV3.Inputs.numColsThe number of columns of the output matrix.BandPart.Inputs.numLower0-D tensor.BatchMatrixBandPart.Inputs.numLowerThe numLower inputMatrixDiag.Inputs.numRowsThe number of rows of the output matrix.MatrixDiagV3.Inputs.numRowsThe number of rows of the output matrix.BandPart.Inputs.numUpper0-D tensor.BatchMatrixBandPart.Inputs.numUpperThe numUpper inputLoadAndRemapMatrix.Inputs.oldTensorNameName of the 2-DTensorto load from checkpoint.MatrixDiag.Inputs.paddingValueThe number to fill the area outside the specified diagonal band with.MatrixDiagPart.Inputs.paddingValueThe value to fill the area outside the specified diagonal band with.MatrixDiagPartV3.Inputs.paddingValueThe value to fill the area outside the specified diagonal band with.MatrixDiagV3.Inputs.paddingValueThe number to fill the area outside the specified diagonal band with.ConjugateTranspose.Inputs.permThe perm inputTranspose.Inputs.permThe perm inputBandedTriangularSolve.Inputs.rhsThe rhs inputBatchMatrixSolve.Inputs.rhsThe rhs inputBatchMatrixSolveLs.Inputs.rhsThe rhs inputBatchMatrixTriangularSolve.Inputs.rhsThe rhs inputMatrixSolveLs.Inputs.rhsShape is[..., M, K].Solve.Inputs.rhsShape is[..., M, K].TriangularSolve.Inputs.rhsShape is[..., M, K].TridiagonalMatMul.Inputs.rhsTensor of shape[..., M, N], representing MxN matrices to the right of multiplication.TridiagonalSolve.Inputs.rhsTensor of shape[..., M, K], representing K right-hand sides per each left-hand side.LoadAndRemapMatrix.Inputs.rowRemappingAn intTensorof row remappings (generally created bygenerate_vocab_remapping).TridiagonalMatMul.Inputs.subdiagTensor of shape[..., 1, M], representing subdiagonals of tri-diagonal matrices to the left of multiplication.TridiagonalMatMul.Inputs.superdiagTensor of shape[..., 1, M], representing superdiagonals of tri-diagonal matrices to the left of multiplication.ConjugateTranspose.Inputs.xThe x inputTranspose.Inputs.xThe x inputFields in org.tensorflow.op.linalg with type parameters of type OperandMethods in org.tensorflow.op.linalg with parameters of type OperandModifier and TypeMethodDescriptionstatic <T extends TType>
BandedTriangularSolve<T> BandedTriangularSolve.create(Scope scope, Operand<T> matrix, Operand<T> rhs, BandedTriangularSolve.Options... options) Factory method to create a class wrapping a new BandedTriangularSolve operation.Factory method to create a class wrapping a new MatrixBandPart operation.static <T extends TNumber>
BatchCholesky<T> Factory method to create a class wrapping a new BatchCholesky operation.static <T extends TNumber>
BatchCholeskyGrad<T> Factory method to create a class wrapping a new BatchCholeskyGrad operation.static <T extends TType>
BatchMatrixBandPart<T> BatchMatrixBandPart.create(Scope scope, Operand<T> input, Operand<TInt64> numLower, Operand<TInt64> numUpper) Factory method to create a class wrapping a new BatchMatrixBandPart operation.static <T extends TType>
BatchMatrixDeterminant<T> Factory method to create a class wrapping a new BatchMatrixDeterminant operation.static <T extends TType>
BatchMatrixDiag<T> Factory method to create a class wrapping a new BatchMatrixDiag operation.static <T extends TType>
BatchMatrixDiagPart<T> Factory method to create a class wrapping a new BatchMatrixDiagPart operation.static <T extends TNumber>
BatchMatrixInverse<T> BatchMatrixInverse.create(Scope scope, Operand<T> input, BatchMatrixInverse.Options... options) Factory method to create a class wrapping a new BatchMatrixInverse operation.static <T extends TType>
BatchMatrixSetDiag<T> Factory method to create a class wrapping a new BatchMatrixSetDiag operation.static <T extends TNumber>
BatchMatrixSolve<T> BatchMatrixSolve.create(Scope scope, Operand<T> matrix, Operand<T> rhs, BatchMatrixSolve.Options... options) Factory method to create a class wrapping a new BatchMatrixSolve operation.static <T extends TNumber>
BatchMatrixSolveLs<T> BatchMatrixSolveLs.create(Scope scope, Operand<T> matrix, Operand<T> rhs, Operand<TFloat64> l2Regularizer, BatchMatrixSolveLs.Options... options) Factory method to create a class wrapping a new BatchMatrixSolveLs operation.static <T extends TNumber>
BatchMatrixTriangularSolve<T> BatchMatrixTriangularSolve.create(Scope scope, Operand<T> matrix, Operand<T> rhs, BatchMatrixTriangularSolve.Options... options) Factory method to create a class wrapping a new BatchMatrixTriangularSolve operation.static <T extends TNumber>
BatchSelfAdjointEig<T> BatchSelfAdjointEig.create(Scope scope, Operand<T> input, BatchSelfAdjointEig.Options... options) Factory method to create a class wrapping a new BatchSelfAdjointEigV2 operation.BatchSvd.create(Scope scope, Operand<T> input, BatchSvd.Options... options) Factory method to create a class wrapping a new BatchSvd operation.Factory method to create a class wrapping a new Cholesky operation.static <T extends TNumber>
CholeskyGrad<T> Factory method to create a class wrapping a new CholeskyGrad operation.static <T extends TType>
ConjugateTranspose<T> Factory method to create a class wrapping a new ConjugateTranspose operation.Factory method to create a class wrapping a new Cross operation.Factory method to create a class wrapping a new MatrixDeterminant operation.Factory method to create a class wrapping a new Eig operation.static <T extends TType>
EuclideanNorm<T> EuclideanNorm.create(Scope scope, Operand<T> input, Operand<? extends TNumber> axis, EuclideanNorm.Options... options) Factory method to create a class wrapping a new EuclideanNorm operation.Inv.create(Scope scope, Operand<T> input, Inv.Options... options) Factory method to create a class wrapping a new MatrixInverse operation.static LoadAndRemapMatrixLoadAndRemapMatrix.create(Scope scope, Operand<TString> ckptPath, Operand<TString> oldTensorName, Operand<TInt64> rowRemapping, Operand<TInt64> colRemapping, Operand<TFloat32> initializingValues, Long numRows, Long numCols, LoadAndRemapMatrix.Options... options) Factory method to create a class wrapping a new LoadAndRemapMatrix operation.static <T extends TType>
LogMatrixDeterminant<T> Factory method to create a class wrapping a new LogMatrixDeterminant operation.Factory method to create a class wrapping a new Lu operation, with the default output types.Factory method to create a class wrapping a new Lu operation.MatMul.create(Scope scope, Operand<T> a, Operand<T> b, MatMul.Options... options) Factory method to create a class wrapping a new MatMul operation.static <T extends TType>
MatrixDiag<T> MatrixDiag.create(Scope scope, Operand<T> diagonal, Operand<TInt32> k, Operand<TInt32> numRows, Operand<TInt32> numCols, Operand<T> paddingValue) Factory method to create a class wrapping a new MatrixDiagV2 operation.static <T extends TType>
MatrixDiagPart<T> Factory method to create a class wrapping a new MatrixDiagPartV2 operation.static <T extends TType>
MatrixDiagPartV3<T> MatrixDiagPartV3.create(Scope scope, Operand<T> input, Operand<TInt32> k, Operand<T> paddingValue, MatrixDiagPartV3.Options... options) Factory method to create a class wrapping a new MatrixDiagPartV3 operation.static <T extends TType>
MatrixDiagV3<T> MatrixDiagV3.create(Scope scope, Operand<T> diagonal, Operand<TInt32> k, Operand<TInt32> numRows, Operand<TInt32> numCols, Operand<T> paddingValue, MatrixDiagV3.Options... options) Factory method to create a class wrapping a new MatrixDiagV3 operation.static <T extends TType>
MatrixExponential<T> Factory method to create a class wrapping a new MatrixExponential operation.static <T extends TType>
MatrixLogarithm<T> Factory method to create a class wrapping a new MatrixLogarithm operation.static <T extends TType>
MatrixSetDiag<T> MatrixSetDiag.create(Scope scope, Operand<T> input, Operand<T> diagonal, Operand<TInt32> k, MatrixSetDiag.Options... options) Factory method to create a class wrapping a new MatrixSetDiagV3 operation.static <T extends TType>
MatrixSolveLs<T> MatrixSolveLs.create(Scope scope, Operand<T> matrix, Operand<T> rhs, Operand<TFloat64> l2Regularizer, MatrixSolveLs.Options... options) Factory method to create a class wrapping a new MatrixSolveLs operation.Qr.create(Scope scope, Operand<T> input, Qr.Options... options) Factory method to create a class wrapping a new Qr operation.static <V extends TNumber, W extends TNumber>
QuantizedMatMul<V> QuantizedMatMul.create(Scope scope, Operand<? extends TNumber> a, Operand<? extends TNumber> b, Operand<TFloat32> minA, Operand<TFloat32> maxA, Operand<TFloat32> minB, Operand<TFloat32> maxB, Class<V> Toutput, Class<W> Tactivation, QuantizedMatMul.Options... options) Factory method to create a class wrapping a new QuantizedMatMul operation.static <W extends TNumber>
QuantizedMatMulWithBias<W> QuantizedMatMulWithBias.create(Scope scope, Operand<? extends TNumber> a, Operand<? extends TNumber> b, Operand<? extends TNumber> bias, Operand<TFloat32> minA, Operand<TFloat32> maxA, Operand<TFloat32> minB, Operand<TFloat32> maxB, Class<W> Toutput, QuantizedMatMulWithBias.Options... options) Factory method to create a class wrapping a new QuantizedMatMulWithBias operation.static <V extends TNumber>
QuantizedMatMulWithBiasAndRelu<V> QuantizedMatMulWithBiasAndRelu.create(Scope scope, Operand<? extends TNumber> a, Operand<? extends TNumber> b, Operand<TFloat32> bias, Operand<TFloat32> minA, Operand<TFloat32> maxA, Operand<TFloat32> minB, Operand<TFloat32> maxB, Class<V> Toutput, QuantizedMatMulWithBiasAndRelu.Options... options) Factory method to create a class wrapping a new QuantizedMatMulWithBiasAndRelu operation.static <W extends TNumber>
QuantizedMatMulWithBiasAndReluAndRequantize<W> QuantizedMatMulWithBiasAndReluAndRequantize.create(Scope scope, Operand<? extends TNumber> a, Operand<? extends TNumber> b, Operand<? extends TNumber> bias, Operand<TFloat32> minA, Operand<TFloat32> maxA, Operand<TFloat32> minB, Operand<TFloat32> maxB, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Class<W> Toutput, QuantizedMatMulWithBiasAndReluAndRequantize.Options... options) Factory method to create a class wrapping a new QuantizedMatMulWithBiasAndReluAndRequantize operation.static <T extends TType>
SelfAdjointEig<T> SelfAdjointEig.create(Scope scope, Operand<T> input, SelfAdjointEig.Options... options) Factory method to create a class wrapping a new SelfAdjointEigV2 operation.Solve.create(Scope scope, Operand<T> matrix, Operand<T> rhs, Solve.Options... options) Factory method to create a class wrapping a new MatrixSolve operation.Factory method to create a class wrapping a new MatrixSquareRoot operation.Svd.create(Scope scope, Operand<T> input, Svd.Options... options) Factory method to create a class wrapping a new Svd operation.static <T extends TType>
TensorDiag<T> Factory method to create a class wrapping a new Diag operation.static <T extends TType>
TensorDiagPart<T> Factory method to create a class wrapping a new DiagPart operation.Factory method to create a class wrapping a new Transpose operation.static <T extends TType>
TriangularSolve<T> TriangularSolve.create(Scope scope, Operand<T> matrix, Operand<T> rhs, TriangularSolve.Options... options) Factory method to create a class wrapping a new MatrixTriangularSolve operation.static <T extends TType>
TridiagonalMatMul<T> TridiagonalMatMul.create(Scope scope, Operand<T> superdiag, Operand<T> maindiag, Operand<T> subdiag, Operand<T> rhs) Factory method to create a class wrapping a new TridiagonalMatMul operation.static <T extends TType>
TridiagonalSolve<T> TridiagonalSolve.create(Scope scope, Operand<T> diagonals, Operand<T> rhs, TridiagonalSolve.Options... options) Factory method to create a class wrapping a new TridiagonalSolve operation.Method parameters in org.tensorflow.op.linalg with type arguments of type Operand -
Uses of Operand in org.tensorflow.op.linalg.sparse
Classes in org.tensorflow.op.linalg.sparse that implement OperandModifier and TypeClassDescriptionfinal classCSRSparseMatrixToDense<T extends TType>Convert a (possibly batched) CSRSparseMatrix to dense.final classConverts a dense tensor to a (possibly batched) CSRSparseMatrix.final classSparse addition of two CSR matrices, C = alpha * A + beta * B.final classSparseMatrixMatMul<T extends TType>Matrix-multiplies a sparse matrix with a dense matrix.final classElement-wise multiplication of a sparse matrix with a dense tensor.final classReturns the number of nonzeroes ofsparse_matrix.final classComputes the Approximate Minimum Degree (AMD) ordering ofinput.final classCalculates the softmax of a CSRSparseMatrix.final classCalculates the gradient of the SparseMatrixSoftmax op.final classComputes the sparse Cholesky decomposition ofinput.final classSparse-matrix-multiplies two CSR matricesaandb.final classTransposes the inner (matrix) dimensions of a CSRSparseMatrix.final classCreates an all-zeros CSRSparseMatrix with shapedense_shape.final classConverts a SparseTensor to a (possibly batched) CSRSparseMatrix.Fields in org.tensorflow.op.linalg.sparse declared as OperandModifier and TypeFieldDescriptionSparseMatrixAdd.Inputs.aA CSRSparseMatrix.SparseMatrixMatMul.Inputs.aA CSRSparseMatrix.SparseMatrixMul.Inputs.aA CSRSparseMatrix.SparseMatrixSparseMatMul.Inputs.aA CSRSparseMatrix.SparseMatrixAdd.Inputs.alphaA constant scalar.SparseMatrixAdd.Inputs.bA CSRSparseMatrix.SparseMatrixMatMul.Inputs.bA dense tensor.SparseMatrixMul.Inputs.bA dense tensor.SparseMatrixSparseMatMul.Inputs.bA CSRSparseMatrix.SparseMatrixAdd.Inputs.betaA constant scalar.CSRSparseMatrixComponents.Inputs.csrSparseMatrixA batched CSRSparseMatrix.DenseToCSRSparseMatrix.Inputs.denseInputA Dense tensor.SparseMatrixZeros.Inputs.denseShapeThe desired matrix shape.SparseTensorToCSRSparseMatrix.Inputs.denseShapeSparseTensor dense shape.SparseMatrixSoftmaxGrad.Inputs.gradSoftmaxThe gradient ofsoftmax.CSRSparseMatrixComponents.Inputs.indexThe index incsr_sparse_matrix's batch.DenseToCSRSparseMatrix.Inputs.indicesIndices of nonzero elements.SparseTensorToCSRSparseMatrix.Inputs.indicesSparseTensor indices.SparseMatrixOrderingAMD.Inputs.inputACSRSparseMatrix.SparseMatrixSparseCholesky.Inputs.inputACSRSparseMatrix.SparseMatrixTranspose.Inputs.inputA CSRSparseMatrix.SparseMatrixSoftmax.Inputs.logitsA CSRSparseMatrix.SparseMatrixSparseCholesky.Inputs.permutationA fill-in reducing permutation matrix.SparseMatrixSoftmaxGrad.Inputs.softmaxA CSRSparseMatrix.CSRSparseMatrixToDense.Inputs.sparseInputA batched CSRSparseMatrix.CSRSparseMatrixToSparseTensor.Inputs.sparseMatrixA (possibly batched) CSRSparseMatrix.SparseMatrixNNZ.Inputs.sparseMatrixA CSRSparseMatrix.SparseTensorToCSRSparseMatrix.Inputs.valuesSparseTensor values.Methods in org.tensorflow.op.linalg.sparse with parameters of type OperandModifier and TypeMethodDescriptionstatic <T extends TType>
CSRSparseMatrixComponents<T> CSRSparseMatrixComponents.create(Scope scope, Operand<? extends TType> csrSparseMatrix, Operand<TInt32> index, Class<T> type) Factory method to create a class wrapping a new CSRSparseMatrixComponents operation.static <T extends TType>
CSRSparseMatrixToDense<T> Factory method to create a class wrapping a new CSRSparseMatrixToDense operation.static <T extends TType>
CSRSparseMatrixToSparseTensor<T> CSRSparseMatrixToSparseTensor.create(Scope scope, Operand<? extends TType> sparseMatrix, Class<T> type) Factory method to create a class wrapping a new CSRSparseMatrixToSparseTensor operation.static DenseToCSRSparseMatrixDenseToCSRSparseMatrix.create(Scope scope, Operand<? extends TType> denseInput, Operand<TInt64> indices) Factory method to create a class wrapping a new DenseToCSRSparseMatrix operation.static <T extends TType>
SparseMatrixAddSparseMatrixAdd.create(Scope scope, Operand<? extends TType> a, Operand<? extends TType> b, Operand<T> alpha, Operand<T> beta) Factory method to create a class wrapping a new SparseMatrixAdd operation.static <T extends TType>
SparseMatrixMatMul<T> SparseMatrixMatMul.create(Scope scope, Operand<? extends TType> a, Operand<T> b, SparseMatrixMatMul.Options... options) Factory method to create a class wrapping a new SparseMatrixMatMul operation.static SparseMatrixMulFactory method to create a class wrapping a new SparseMatrixMul operation.static SparseMatrixNNZFactory method to create a class wrapping a new SparseMatrixNNZ operation.static SparseMatrixOrderingAMDFactory method to create a class wrapping a new SparseMatrixOrderingAMD operation.static <T extends TNumber>
SparseMatrixSoftmaxFactory method to create a class wrapping a new SparseMatrixSoftmax operation.static <T extends TNumber>
SparseMatrixSoftmaxGradSparseMatrixSoftmaxGrad.create(Scope scope, Operand<? extends TType> softmax, Operand<? extends TType> gradSoftmax, Class<T> type) Factory method to create a class wrapping a new SparseMatrixSoftmaxGrad operation.static <T extends TType>
SparseMatrixSparseCholeskySparseMatrixSparseCholesky.create(Scope scope, Operand<? extends TType> input, Operand<TInt32> permutation, Class<T> type) Factory method to create a class wrapping a new SparseMatrixSparseCholesky operation.static <T extends TType>
SparseMatrixSparseMatMulSparseMatrixSparseMatMul.create(Scope scope, Operand<? extends TType> a, Operand<? extends TType> b, Class<T> type, SparseMatrixSparseMatMul.Options... options) Factory method to create a class wrapping a new SparseMatrixSparseMatMul operation.static <T extends TType>
SparseMatrixTransposeSparseMatrixTranspose.create(Scope scope, Operand<? extends TType> input, Class<T> type, SparseMatrixTranspose.Options... options) Factory method to create a class wrapping a new SparseMatrixTranspose operation.static <T extends TType>
SparseMatrixZerosFactory method to create a class wrapping a new SparseMatrixZeros operation.SparseTensorToCSRSparseMatrix.create(Scope scope, Operand<TInt64> indices, Operand<? extends TType> values, Operand<TInt64> denseShape) Factory method to create a class wrapping a new SparseTensorToCSRSparseMatrix operation. -
Uses of Operand in org.tensorflow.op.math
Classes in org.tensorflow.op.math that implement OperandModifier and TypeClassDescriptionfinal classComputes the absolute value of a tensor.final classAccumulateN<T extends TType>Returns the element-wise sum of a list of tensors.final classComputes acos of x element-wise.final classComputes inverse hyperbolic cosine of x element-wise.final classReturns x + y element-wise.final classAdd all input tensors element wise.final classReturns the argument of a complex number.final classReturns the truth value of abs(x-y) < tolerance element-wise.final classReturns the index with the largest value across dimensions of a tensor.final classReturns the index with the smallest value across dimensions of a tensor.final classComputes the trignometric inverse sine of x element-wise.final classComputes inverse hyperbolic sine of x element-wise.final classComputes the trignometric inverse tangent of x element-wise.final classComputes arctangent ofy/xelement-wise, respecting signs of the arguments.final classComputes inverse hyperbolic tangent of x element-wise.final classThe BesselI0 operationfinal classThe BesselI0e operationfinal classThe BesselI1 operationfinal classThe BesselI1e operationfinal classCompute the regularized incomplete beta integral \(I_x(a, b)\).final classCounts the number of occurrences of each value in an integer array.final classReturns element-wise smallest integer not less than x.final classComplexAbs<U extends TNumber>Computes the complex absolute value of a tensor.final classReturns the complex conjugate of a complex number.final classComputes cos of x element-wise.final classComputes hyperbolic cosine of x element-wise.final classCompute the cumulative product of the tensorxalongaxis.final classCompute the cumulative sum of the tensorxalongaxis.final classCumulativeLogsumexp<T extends TNumber>Compute the cumulative product of the tensorxalongaxis.final classDenseBincount<U extends TNumber>Counts the number of occurrences of each value in an integer array.final classComputes Psi, the derivative of Lgamma (the log of the absolute value ofGamma(x)), element-wise.final classReturns x / y element-wise.final classReturns 0 if the denominator is zero.final classReturns the truth value of (x == y) element-wise.final classComputes the Gauss error function ofxelement-wise.final classComputes the complementary error function ofxelement-wise.final classThe Erfinv operationfinal classComputes exponential of x element-wise.final classComputesexp(x) - 1element-wise. i.e.final classOutput a fact about factorials.final classReturns element-wise largest integer not greater than x.final classReturns x // y element-wise.final classReturns element-wise remainder of division.final classReturns the truth value of (x > y) element-wise.final classReturns the truth value of (x >= y) element-wise.final classCompute the lower regularized incomplete Gamma functionP(a, x).final classCompute the upper regularized incomplete Gamma functionQ(a, x).final classIgammaGradA<T extends TNumber>Computes the gradient ofigamma(a, x)wrta.final classReturns the imaginary part of a complex number.final classInvertPermutation<T extends TNumber>Computes the inverse permutation of a tensor.final classReturns which elements of x are finite.final classReturns which elements of x are Inf.final classReturns which elements of x are NaN.final classReturns the truth value of (x < y) element-wise.final classReturns the truth value of (x <= y) element-wise.final classComputes the log of the absolute value ofGamma(x)element-wise.final classComputes natural logarithm of x element-wise.final classComputes natural logarithm of (1 + x) element-wise.final classReturns the truth value of x AND y element-wise.final classReturns the truth value ofNOT xelement-wise.final classReturns the truth value of x OR y element-wise.final classReturns the max of x and y (i.e. x > y ?final classComputes the mean of elements across dimensions of a tensor.final classReturns the min of x and y (i.e. x < y ?final classReturns element-wise remainder of division.final classReturns x * y element-wise.final classReturns x * y element-wise.final classThe Ndtri operationfinal classComputes numerical negative value element-wise.final classReturns the next representable value ofx1in the direction ofx2, element-wise.final classReturns the truth value of (x !final classCompute the polygamma function \(\psi^{(n)}(x)\).final classComputes element-wise population count (a.k.a. popcount, bitsum, bitcount).final classComputes the power of one value to another.final classReturns the real part of a complex number.final classReturns x / y element-wise for real types.final classReciprocal<T extends TType>Computes the reciprocal of x element-wise.final classReciprocalGrad<T extends TType>Computes the gradient for the inverse ofxwrt its input.final classReturns element-wise integer closest to x.final classRounds the values of a tensor to the nearest integer, element-wise.final classComputes reciprocal of square root of x element-wise.final classComputes the gradient for the rsqrt ofxwrt its input.final classSegmentMax<T extends TNumber>Computes the maximum along segments of a tensor.final classSegmentMean<T extends TType>Computes the mean along segments of a tensor.final classSegmentMin<T extends TNumber>Computes the minimum along segments of a tensor.final classSegmentProd<T extends TType>Computes the product along segments of a tensor.final classSegmentSum<T extends TType>Computes the sum along segments of a tensor.final classComputes sigmoid ofxelement-wise.final classSigmoidGrad<T extends TType>Computes the gradient of the sigmoid ofxwrt its input.final classReturns an element-wise indication of the sign of a number.final classComputes sine of x element-wise.final classComputes hyperbolic sine of x element-wise.final classSobolSample<T extends TNumber>Generates points from the Sobol sequence.final classThe Softplus operationfinal classSoftplusGrad<T extends TNumber>Computes softplus gradients for a softplus operation.final classComputes square root of x element-wise.final classComputes the gradient for the sqrt ofxwrt its input.final classComputes square of x element-wise.final classSquaredDifference<T extends TType>Returns conj(x - y)(x - y) element-wise.final classReturns x - y element-wise.final classComputes tan of x element-wise.final classComputes hyperbolic tangent ofxelement-wise.final classComputes the gradient for the tanh ofxwrt its input.final classTruncateDiv<T extends TType>Returns x / y element-wise, rounded towards zero.final classTruncateMod<T extends TNumber>Returns element-wise remainder of division.final classUniformQuantizedAdd<T extends TNumber>Perform quantized add of quantized Tensorlhsand quantized Tensorrhsto make quantizedoutput.final classUnsortedSegmentMax<T extends TNumber>Computes the maximum along segments of a tensor.final classUnsortedSegmentMin<T extends TNumber>Computes the minimum along segments of a tensor.final classUnsortedSegmentProd<T extends TType>Computes the product along segments of a tensor.final classUnsortedSegmentSum<T extends TType>Computes the sum along segments of a tensor.final classReturns 0 if x == 0, and x / y otherwise, elementwise.final classReturns 0 if x == 0, and x * log1p(y) otherwise, elementwise.final classReturns 0 if x == 0, and x * log(y) otherwise, elementwise.final classCompute the Hurwitz zeta function \(\zeta(x, q)\).Fields in org.tensorflow.op.math declared as OperandModifier and TypeFieldDescriptionBetainc.Inputs.aThe a inputIgamma.Inputs.aThe a inputIgammac.Inputs.aThe a inputIgammaGradA.Inputs.aThe a inputPolygamma.Inputs.aThe a inputBincount.Inputs.arrint32Tensor.Cumprod.Inputs.axisATensorof typeint32(default: 0).Cumsum.Inputs.axisATensorof typeint32(default: 0).CumulativeLogsumexp.Inputs.axisATensorof typeint32(default: 0).Mean.Inputs.axisThe dimensions to reduce.Betainc.Inputs.bThe b inputSegmentMax.Inputs.dataThe data inputSegmentMean.Inputs.dataThe data inputSegmentMin.Inputs.dataThe data inputSegmentProd.Inputs.dataThe data inputSegmentSum.Inputs.dataThe data inputUnsortedSegmentMax.Inputs.dataThe data inputUnsortedSegmentMin.Inputs.dataThe data inputUnsortedSegmentProd.Inputs.dataThe data inputUnsortedSegmentSum.Inputs.dataThe data inputSobolSample.Inputs.dimPositive scalarTensorrepresenting each sample's dimension.ArgMax.Inputs.dimensionint16, int32 or int64, must be in the range[-rank(input), rank(input)).ArgMin.Inputs.dimensionint32 or int64, must be in the range[-rank(input), rank(input)).ReciprocalGrad.Inputs.dyThe dy inputRsqrtGrad.Inputs.dyThe dy inputSigmoidGrad.Inputs.dyThe dy inputSqrtGrad.Inputs.dyThe dy inputTanhGrad.Inputs.dyThe dy inputSoftplus.Inputs.featuresThe features inputSoftplusGrad.Inputs.featuresThe features passed as input to the corresponding softplus operation.SoftplusGrad.Inputs.gradientsThe backpropagated gradients to the corresponding softplus operation.Angle.Inputs.inputThe input inputArgMax.Inputs.inputThe input inputArgMin.Inputs.inputThe input inputConj.Inputs.inputThe input inputDenseBincount.Inputs.input1D or 2D intTensor.Imag.Inputs.inputThe input inputMean.Inputs.inputThe tensor to reduce.Real.Inputs.inputThe input inputRequantizationRangePerChannel.Inputs.inputThe original input tensor.RequantizePerChannel.Inputs.inputThe original input tensor.RequantizationRangePerChannel.Inputs.inputMaxThe maximum value of the input tensor.RequantizePerChannel.Inputs.inputMaxThe maximum value of the input tensor.RequantizationRangePerChannel.Inputs.inputMinThe minimum value of the input tensorRequantizePerChannel.Inputs.inputMinThe minimum value of the input tensorUniformQuantizedAdd.Inputs.lhsMust be a quantized tensor.UniformQuantizedAdd.Inputs.lhsScalesThe float value(s) used as scale factors when quantizing the original data thatlhsrepresents.UniformQuantizedAdd.Inputs.lhsZeroPointsThe int32 value(s) used as zero points when quantizing original data thatlhsrepresents.QuantizedAdd.Inputs.maxXThe float value that the highest quantizedxvalue represents.QuantizedMul.Inputs.maxXThe float value that the highest quantizedxvalue represents.QuantizedAdd.Inputs.maxYThe float value that the highest quantizedyvalue represents.QuantizedMul.Inputs.maxYThe float value that the highest quantizedyvalue represents.QuantizedAdd.Inputs.minXThe float value that the lowest quantizedxvalue represents.QuantizedMul.Inputs.minXThe float value that the lowest quantizedxvalue represents.QuantizedAdd.Inputs.minYThe float value that the lowest quantizedyvalue represents.QuantizedMul.Inputs.minYThe float value that the lowest quantizedyvalue represents.SobolSample.Inputs.numResultsPositive scalarTensorof dtype int32.SegmentMax.Inputs.numSegmentsThe numSegments inputSegmentMin.Inputs.numSegmentsThe numSegments inputSegmentProd.Inputs.numSegmentsThe numSegments inputSegmentSum.Inputs.numSegmentsThe numSegments inputUnsortedSegmentMax.Inputs.numSegmentsThe numSegments inputUnsortedSegmentMin.Inputs.numSegmentsThe numSegments inputUnsortedSegmentProd.Inputs.numSegmentsThe numSegments inputUnsortedSegmentSum.Inputs.numSegmentsThe numSegments inputUniformQuantizedAdd.Inputs.outputScalesThe float value(s) to use as scale factors when quantizing original data thatoutputrepresents.UniformQuantizedAdd.Inputs.outputZeroPointsThe int32 value(s) used as zero points when quantizing original data that output represents.Zeta.Inputs.qThe q inputRequantizePerChannel.Inputs.requestedOutputMaxThe maximum value of the output tensor requested.RequantizePerChannel.Inputs.requestedOutputMinThe minimum value of the output tensor requested.UniformQuantizedAdd.Inputs.rhsMust be a quantized tensor.UniformQuantizedAdd.Inputs.rhsScalesThe float value(s) used as scale factors when quantizing the original data thatrhsrepresents.UniformQuantizedAdd.Inputs.rhsZeroPointsThe int32 value(s) used as zero points when quantizing original data thatrhsrepresents.SegmentMax.Inputs.segmentIdsA 1-D tensor whose size is equal to the size ofdata's first dimension.SegmentMean.Inputs.segmentIdsA 1-D tensor whose size is equal to the size ofdata's first dimension.SegmentMin.Inputs.segmentIdsA 1-D tensor whose size is equal to the size ofdata's first dimension.SegmentProd.Inputs.segmentIdsA 1-D tensor whose size is equal to the size ofdata's first dimension.SegmentSum.Inputs.segmentIdsA 1-D tensor whose size is equal to the size ofdata's first dimension.UnsortedSegmentMax.Inputs.segmentIdsA tensor whose shape is a prefix ofdata.shape.UnsortedSegmentMin.Inputs.segmentIdsA tensor whose shape is a prefix ofdata.shape.UnsortedSegmentProd.Inputs.segmentIdsA tensor whose shape is a prefix ofdata.shape.UnsortedSegmentSum.Inputs.segmentIdsA tensor whose shape is a prefix ofdata.shape.Bincount.Inputs.sizeOutputnon-negative int32 scalarTensor.DenseBincount.Inputs.sizeOutputnon-negative int scalarTensor.SobolSample.Inputs.skipPositive scalarTensorof dtype int32.Bincount.Inputs.weightsis an int32, int64, float32, or float64Tensorwith the same shape asarr, or a length-0Tensor, in which case it acts as all weights equal to 1.DenseBincount.Inputs.weightsis an int32, int64, float32, or float64Tensorwith the same shape asarr, or a length-0Tensor, in which case it acts as all weights equal to 1.Abs.Inputs.xThe x inputAcos.Inputs.xThe x inputAcosh.Inputs.xThe x inputAdd.Inputs.xThe x inputApproximateEqual.Inputs.xThe x inputAsin.Inputs.xThe x inputAsinh.Inputs.xThe x inputAtan.Inputs.xThe x inputAtan2.Inputs.xThe x inputAtanh.Inputs.xThe x inputBesselI0.Inputs.xThe x inputBesselI0e.Inputs.xThe x inputBesselI1.Inputs.xThe x inputBesselI1e.Inputs.xThe x inputBetainc.Inputs.xThe x inputCeil.Inputs.xThe x inputComplexAbs.Inputs.xThe x inputCos.Inputs.xThe x inputCosh.Inputs.xThe x inputCumprod.Inputs.xATensor.Cumsum.Inputs.xATensor.CumulativeLogsumexp.Inputs.xATensor.Digamma.Inputs.xThe x inputDiv.Inputs.xThe x inputDivNoNan.Inputs.xThe x inputEqual.Inputs.xThe x inputErf.Inputs.xThe x inputErfc.Inputs.xThe x inputerfinv.Inputs.xThe x inputExp.Inputs.xThe x inputExpm1.Inputs.xThe x inputFloor.Inputs.xThe x inputFloorDiv.Inputs.xThe x inputFloorMod.Inputs.xThe x inputGreater.Inputs.xThe x inputGreaterEqual.Inputs.xThe x inputIgamma.Inputs.xThe x inputIgammac.Inputs.xThe x inputIgammaGradA.Inputs.xThe x inputInvertPermutation.Inputs.x1-D.IsFinite.Inputs.xThe x inputIsInf.Inputs.xThe x inputIsNan.Inputs.xThe x inputLess.Inputs.xThe x inputLessEqual.Inputs.xThe x inputLgamma.Inputs.xThe x inputLog.Inputs.xThe x inputLog1p.Inputs.xThe x inputLogicalAnd.Inputs.xThe x inputLogicalNot.Inputs.xATensorof typebool.LogicalOr.Inputs.xThe x inputMaximum.Inputs.xThe x inputMinimum.Inputs.xThe x inputMod.Inputs.xThe x inputMul.Inputs.xThe x inputMulNoNan.Inputs.xThe x inputNdtri.Inputs.xThe x inputNeg.Inputs.xThe x inputNotEqual.Inputs.xThe x inputPolygamma.Inputs.xThe x inputPopulationCount.Inputs.xThe x inputPow.Inputs.xThe x inputQuantizedAdd.Inputs.xThe x inputQuantizedMul.Inputs.xThe x inputRealDiv.Inputs.xThe x inputReciprocal.Inputs.xThe x inputRint.Inputs.xThe x inputRound.Inputs.xThe x inputRsqrt.Inputs.xThe x inputSigmoid.Inputs.xThe x inputSign.Inputs.xThe x inputSin.Inputs.xThe x inputSinh.Inputs.xThe x inputSqrt.Inputs.xThe x inputSquare.Inputs.xThe x inputSquaredDifference.Inputs.xThe x inputSub.Inputs.xThe x inputTan.Inputs.xThe x inputTanh.Inputs.xThe x inputTruncateDiv.Inputs.xThe x inputTruncateMod.Inputs.xThe x inputXdivy.Inputs.xThe x inputXlog1py.Inputs.xThe x inputXlogy.Inputs.xThe x inputZeta.Inputs.xThe x inputNextAfter.Inputs.x1The x1 inputNextAfter.Inputs.x2The x2 inputAdd.Inputs.yThe y inputApproximateEqual.Inputs.yThe y inputAtan2.Inputs.yThe y inputDiv.Inputs.yThe y inputDivNoNan.Inputs.yThe y inputEqual.Inputs.yThe y inputFloorDiv.Inputs.yThe y inputFloorMod.Inputs.yThe y inputGreater.Inputs.yThe y inputGreaterEqual.Inputs.yThe y inputLess.Inputs.yThe y inputLessEqual.Inputs.yThe y inputLogicalAnd.Inputs.yThe y inputLogicalOr.Inputs.yThe y inputMaximum.Inputs.yThe y inputMinimum.Inputs.yThe y inputMod.Inputs.yThe y inputMul.Inputs.yThe y inputMulNoNan.Inputs.yThe y inputNotEqual.Inputs.yThe y inputPow.Inputs.yThe y inputQuantizedAdd.Inputs.yThe y inputQuantizedMul.Inputs.yThe y inputRealDiv.Inputs.yThe y inputReciprocalGrad.Inputs.yThe y inputRsqrtGrad.Inputs.yThe y inputSigmoidGrad.Inputs.yThe y inputSqrtGrad.Inputs.yThe y inputSquaredDifference.Inputs.yThe y inputSub.Inputs.yThe y inputTanhGrad.Inputs.yThe y inputTruncateDiv.Inputs.yThe y inputTruncateMod.Inputs.yThe y inputXdivy.Inputs.yThe y inputXlog1py.Inputs.yThe y inputXlogy.Inputs.yThe y inputFields in org.tensorflow.op.math with type parameters of type OperandModifier and TypeFieldDescriptionAccumulateN.Inputs.inputsA list ofTensorobjects, each with same shape and type.AddN.Inputs.inputsThe inputs inputMethods in org.tensorflow.op.math with parameters of type OperandModifier and TypeMethodDescriptionFactory method to create a class wrapping a new Abs operation.Factory method to create a class wrapping a new Acos operation.Factory method to create a class wrapping a new Acosh operation.Factory method to create a class wrapping a new Add operation.Factory method to create a class wrapping a new Angle operation, with the default output types.Factory method to create a class wrapping a new Angle operation.static <T extends TType>
ApproximateEqualApproximateEqual.create(Scope scope, Operand<T> x, Operand<T> y, ApproximateEqual.Options... options) Factory method to create a class wrapping a new ApproximateEqual operation.Factory method to create a class wrapping a new ArgMax operation, with the default output types.ArgMax.create(Scope scope, Operand<? extends TType> input, Operand<? extends TNumber> dimension, Class<V> outputType) Factory method to create a class wrapping a new ArgMax operation.Factory method to create a class wrapping a new ArgMin operation, with the default output types.ArgMin.create(Scope scope, Operand<? extends TType> input, Operand<? extends TNumber> dimension, Class<V> outputType) Factory method to create a class wrapping a new ArgMin operation.Factory method to create a class wrapping a new Asin operation.Factory method to create a class wrapping a new Asinh operation.Factory method to create a class wrapping a new Atan operation.Factory method to create a class wrapping a new Atan2 operation.Factory method to create a class wrapping a new Atanh operation.Factory method to create a class wrapping a new BesselI0 operation.Factory method to create a class wrapping a new BesselI0e operation.Factory method to create a class wrapping a new BesselI1 operation.Factory method to create a class wrapping a new BesselI1e operation.Factory method to create a class wrapping a new Betainc operation.Factory method to create a class wrapping a new Bincount operation.Factory method to create a class wrapping a new Ceil operation.static ComplexAbs<TFloat32> Factory method to create a class wrapping a new ComplexAbs operation, with the default output types.static <U extends TNumber>
ComplexAbs<U> Factory method to create a class wrapping a new ComplexAbs operation.Factory method to create a class wrapping a new Conj operation.Factory method to create a class wrapping a new Cos operation.Factory method to create a class wrapping a new Cosh operation.Cumprod.create(Scope scope, Operand<T> x, Operand<? extends TNumber> axis, Cumprod.Options... options) Factory method to create a class wrapping a new Cumprod operation.Cumsum.create(Scope scope, Operand<T> x, Operand<? extends TNumber> axis, Cumsum.Options... options) Factory method to create a class wrapping a new Cumsum operation.static <T extends TNumber>
CumulativeLogsumexp<T> CumulativeLogsumexp.create(Scope scope, Operand<T> x, Operand<? extends TNumber> axis, CumulativeLogsumexp.Options... options) Factory method to create a class wrapping a new CumulativeLogsumexp operation.static <U extends TNumber, T extends TNumber>
DenseBincount<U> DenseBincount.create(Scope scope, Operand<T> input, Operand<T> sizeOutput, Operand<U> weights, DenseBincount.Options... options) Factory method to create a class wrapping a new DenseBincount operation.Factory method to create a class wrapping a new Digamma operation.Factory method to create a class wrapping a new Div operation.Factory method to create a class wrapping a new DivNoNan operation.Equal.create(Scope scope, Operand<T> x, Operand<T> y, Equal.Options... options) Factory method to create a class wrapping a new Equal operation.Factory method to create a class wrapping a new Erf operation.Factory method to create a class wrapping a new Erfc operation.Factory method to create a class wrapping a new Erfinv operation.Factory method to create a class wrapping a new Exp operation.Factory method to create a class wrapping a new Expm1 operation.Factory method to create a class wrapping a new Floor operation.Factory method to create a class wrapping a new FloorDiv operation.Factory method to create a class wrapping a new FloorMod operation.Factory method to create a class wrapping a new Greater operation.static <T extends TNumber>
GreaterEqualFactory method to create a class wrapping a new GreaterEqual operation.Factory method to create a class wrapping a new Igamma operation.Factory method to create a class wrapping a new Igammac operation.static <T extends TNumber>
IgammaGradA<T> Factory method to create a class wrapping a new IgammaGradA operation.Factory method to create a class wrapping a new Imag operation, with the default output types.Factory method to create a class wrapping a new Imag operation.static <T extends TNumber>
InvertPermutation<T> Factory method to create a class wrapping a new InvertPermutation operation.static IsFiniteFactory method to create a class wrapping a new IsFinite operation.static IsInfFactory method to create a class wrapping a new IsInf operation.static IsNanFactory method to create a class wrapping a new IsNan operation.Factory method to create a class wrapping a new Less operation.Factory method to create a class wrapping a new LessEqual operation.Factory method to create a class wrapping a new Lgamma operation.Factory method to create a class wrapping a new Log operation.Factory method to create a class wrapping a new Log1p operation.static LogicalAndFactory method to create a class wrapping a new LogicalAnd operation.static LogicalNotFactory method to create a class wrapping a new LogicalNot operation.static LogicalOrFactory method to create a class wrapping a new LogicalOr operation.Factory method to create a class wrapping a new Maximum operation.Mean.create(Scope scope, Operand<T> input, Operand<? extends TNumber> axis, Mean.Options... options) Factory method to create a class wrapping a new Mean operation.Factory method to create a class wrapping a new Minimum operation.Factory method to create a class wrapping a new Mod operation.Factory method to create a class wrapping a new Mul operation.Factory method to create a class wrapping a new MulNoNan operation.Factory method to create a class wrapping a new Ndtri operation.Factory method to create a class wrapping a new Neg operation.Factory method to create a class wrapping a new NextAfter operation.NotEqual.create(Scope scope, Operand<T> x, Operand<T> y, NotEqual.Options... options) Factory method to create a class wrapping a new NotEqual operation.Factory method to create a class wrapping a new Polygamma operation.static PopulationCountFactory method to create a class wrapping a new PopulationCount operation.Factory method to create a class wrapping a new Pow operation.static <V extends TNumber>
QuantizedAdd<V> QuantizedAdd.create(Scope scope, Operand<? extends TNumber> x, Operand<? extends TNumber> y, Operand<TFloat32> minX, Operand<TFloat32> maxX, Operand<TFloat32> minY, Operand<TFloat32> maxY, Class<V> Toutput) Factory method to create a class wrapping a new QuantizedAdd operation.static <V extends TNumber>
QuantizedMul<V> QuantizedMul.create(Scope scope, Operand<? extends TNumber> x, Operand<? extends TNumber> y, Operand<TFloat32> minX, Operand<TFloat32> maxX, Operand<TFloat32> minY, Operand<TFloat32> maxY, Class<V> Toutput) Factory method to create a class wrapping a new QuantizedMul operation.Factory method to create a class wrapping a new Real operation, with the default output types.Factory method to create a class wrapping a new Real operation.Factory method to create a class wrapping a new RealDiv operation.static <T extends TType>
Reciprocal<T> Factory method to create a class wrapping a new Reciprocal operation.static <T extends TType>
ReciprocalGrad<T> Factory method to create a class wrapping a new ReciprocalGrad operation.RequantizationRangePerChannel.create(Scope scope, Operand<? extends TNumber> input, Operand<TFloat32> inputMin, Operand<TFloat32> inputMax, Float clipValueMax) Factory method to create a class wrapping a new RequantizationRangePerChannel operation.static <U extends TNumber>
RequantizePerChannel<U> RequantizePerChannel.create(Scope scope, Operand<? extends TNumber> input, Operand<TFloat32> inputMin, Operand<TFloat32> inputMax, Operand<TFloat32> requestedOutputMin, Operand<TFloat32> requestedOutputMax, Class<U> outType) Factory method to create a class wrapping a new RequantizePerChannel operation.Factory method to create a class wrapping a new Rint operation.Factory method to create a class wrapping a new Round operation.Factory method to create a class wrapping a new Rsqrt operation.Factory method to create a class wrapping a new RsqrtGrad operation.static <T extends TNumber>
SegmentMax<T> SegmentMax.create(Scope scope, Operand<T> data, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments) Factory method to create a class wrapping a new SegmentMaxV2 operation.static <T extends TType>
SegmentMean<T> Factory method to create a class wrapping a new SegmentMean operation.static <T extends TNumber>
SegmentMin<T> SegmentMin.create(Scope scope, Operand<T> data, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments) Factory method to create a class wrapping a new SegmentMinV2 operation.static <T extends TType>
SegmentProd<T> SegmentProd.create(Scope scope, Operand<T> data, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments) Factory method to create a class wrapping a new SegmentProdV2 operation.static <T extends TType>
SegmentSum<T> SegmentSum.create(Scope scope, Operand<T> data, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments) Factory method to create a class wrapping a new SegmentSumV2 operation.Factory method to create a class wrapping a new Sigmoid operation.static <T extends TType>
SigmoidGrad<T> Factory method to create a class wrapping a new SigmoidGrad operation.Factory method to create a class wrapping a new Sign operation.Factory method to create a class wrapping a new Sin operation.Factory method to create a class wrapping a new Sinh operation.static SobolSample<TFloat32> SobolSample.create(Scope scope, Operand<TInt32> dim, Operand<TInt32> numResults, Operand<TInt32> skip) Factory method to create a class wrapping a new SobolSample operation, with the default output types.static <T extends TNumber>
SobolSample<T> SobolSample.create(Scope scope, Operand<TInt32> dim, Operand<TInt32> numResults, Operand<TInt32> skip, Class<T> dtype) Factory method to create a class wrapping a new SobolSample operation.Factory method to create a class wrapping a new Softplus operation.static <T extends TNumber>
SoftplusGrad<T> Factory method to create a class wrapping a new SoftplusGrad operation.Factory method to create a class wrapping a new Sqrt operation.Factory method to create a class wrapping a new SqrtGrad operation.Factory method to create a class wrapping a new Square operation.static <T extends TType>
SquaredDifference<T> Factory method to create a class wrapping a new SquaredDifference operation.Factory method to create a class wrapping a new Sub operation.Factory method to create a class wrapping a new Tan operation.Factory method to create a class wrapping a new Tanh operation.Factory method to create a class wrapping a new TanhGrad operation.static <T extends TType>
TruncateDiv<T> Factory method to create a class wrapping a new TruncateDiv operation.static <T extends TNumber>
TruncateMod<T> Factory method to create a class wrapping a new TruncateMod operation.static <T extends TNumber>
UniformQuantizedAdd<T> UniformQuantizedAdd.create(Scope scope, Operand<T> lhs, Operand<T> rhs, Operand<TFloat32> lhsScales, Operand<TInt32> lhsZeroPoints, Operand<TFloat32> rhsScales, Operand<TInt32> rhsZeroPoints, Operand<TFloat32> outputScales, Operand<TInt32> outputZeroPoints, Long lhsQuantizationMinVal, Long lhsQuantizationMaxVal, Long rhsQuantizationMinVal, Long rhsQuantizationMaxVal, Long outputQuantizationMinVal, Long outputQuantizationMaxVal, UniformQuantizedAdd.Options... options) Factory method to create a class wrapping a new UniformQuantizedAdd operation.static <T extends TNumber>
UnsortedSegmentMax<T> UnsortedSegmentMax.create(Scope scope, Operand<T> data, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments) Factory method to create a class wrapping a new UnsortedSegmentMax operation.static <T extends TNumber>
UnsortedSegmentMin<T> UnsortedSegmentMin.create(Scope scope, Operand<T> data, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments) Factory method to create a class wrapping a new UnsortedSegmentMin operation.static <T extends TType>
UnsortedSegmentProd<T> UnsortedSegmentProd.create(Scope scope, Operand<T> data, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments) Factory method to create a class wrapping a new UnsortedSegmentProd operation.static <T extends TType>
UnsortedSegmentSum<T> UnsortedSegmentSum.create(Scope scope, Operand<T> data, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments) Factory method to create a class wrapping a new UnsortedSegmentSum operation.Factory method to create a class wrapping a new Xdivy operation.Factory method to create a class wrapping a new Xlog1py operation.Factory method to create a class wrapping a new Xlogy operation.Factory method to create a class wrapping a new Zeta operation.Method parameters in org.tensorflow.op.math with type arguments of type OperandModifier and TypeMethodDescriptionstatic <T extends TType>
AccumulateN<T> Factory method to create a class wrapping a new AccumulateNV2 operation.Factory method to create a class wrapping a new AddN operation. -
Uses of Operand in org.tensorflow.op.math.special
Classes in org.tensorflow.op.math.special that implement OperandModifier and TypeClassDescriptionfinal classThe BesselJ0 operationfinal classThe BesselJ1 operationfinal classThe BesselK0 operationfinal classThe BesselK0e operationfinal classThe BesselK1 operationfinal classThe BesselK1e operationfinal classThe BesselY0 operationfinal classThe BesselY1 operationfinal classThe Dawsn operationfinal classThe Expint operationfinal classFresnelCos<T extends TNumber>The FresnelCos operationfinal classFresnelSin<T extends TNumber>The FresnelSin operationfinal classThe Spence operationFields in org.tensorflow.op.math.special declared as OperandModifier and TypeFieldDescriptionBesselJ0.Inputs.xThe x inputBesselJ1.Inputs.xThe x inputBesselK0.Inputs.xThe x inputBesselK0e.Inputs.xThe x inputBesselK1.Inputs.xThe x inputBesselK1e.Inputs.xThe x inputBesselY0.Inputs.xThe x inputBesselY1.Inputs.xThe x inputDawsn.Inputs.xThe x inputExpint.Inputs.xThe x inputFresnelCos.Inputs.xThe x inputFresnelSin.Inputs.xThe x inputSpence.Inputs.xThe x inputMethods in org.tensorflow.op.math.special with parameters of type OperandModifier and TypeMethodDescriptionFactory method to create a class wrapping a new BesselJ0 operation.Factory method to create a class wrapping a new BesselJ1 operation.Factory method to create a class wrapping a new BesselK0 operation.Factory method to create a class wrapping a new BesselK0e operation.Factory method to create a class wrapping a new BesselK1 operation.Factory method to create a class wrapping a new BesselK1e operation.Factory method to create a class wrapping a new BesselY0 operation.Factory method to create a class wrapping a new BesselY1 operation.Factory method to create a class wrapping a new Dawsn operation.Factory method to create a class wrapping a new Expint operation.static <T extends TNumber>
FresnelCos<T> Factory method to create a class wrapping a new FresnelCos operation.static <T extends TNumber>
FresnelSin<T> Factory method to create a class wrapping a new FresnelSin operation.Factory method to create a class wrapping a new Spence operation. -
Uses of Operand in org.tensorflow.op.nn
Classes in org.tensorflow.op.nn that implement OperandModifier and TypeClassDescriptionfinal classPerforms average pooling on the input.final classPerforms 3D average pooling on the input.final classAvgPool3dGrad<T extends TNumber>Computes gradients of average pooling function.final classAvgPoolGrad<T extends TNumber>Computes gradients of the average pooling function.final classBatchNormWithGlobalNormalization<T extends TType>Batch normalization.final classAddsbiastovalue.final classBiasAddGrad<T extends TType>The backward operation for "BiasAdd" on the "bias" tensor.final classComputes a N-D convolution given (N+1+batch_dims)-Dinputand (N+2)-Dfiltertensors.final classComputes a 2-D convolution given 4-Dinputandfiltertensors.final classConv2dBackpropFilter<T extends TNumber>Computes the gradients of convolution with respect to the filter.final classConv2dBackpropFilterV2<T extends TNumber>Computes the gradients of convolution with respect to the filter.final classConv2dBackpropInput<T extends TNumber>Computes the gradients of convolution with respect to the input.final classConv2dBackpropInputV2<T extends TNumber>Computes the gradients of convolution with respect to the input.final classComputes a 3-D convolution given 5-Dinputandfiltertensors.final classConv3dBackpropFilter<T extends TNumber>Computes the gradients of 3-D convolution with respect to the filter.final classConv3dBackpropInput<U extends TNumber>Computes the gradients of 3-D convolution with respect to the input.final classCudnnRNNCanonicalToParams<T extends TNumber>Converts CudnnRNN params from canonical form to usable form.final classCudnnRnnParamsSize<T extends TNumber>Computes size of weights that can be used by a Cudnn RNN model.final classDataFormatDimMap<T extends TNumber>Returns the dimension index in the destination data format given the one in the source data format.final classDataFormatVecPermute<T extends TNumber>Permute input tensor fromsrc_formattodst_format.final classDepthToSpace<T extends TType>DepthToSpace for tensors of type T.final classDepthwiseConv2dNative<T extends TNumber>Computes a 2-D depthwise convolution given 4-Dinputandfiltertensors.final classDepthwiseConv2dNativeBackpropFilter<T extends TNumber>Computes the gradients of depthwise convolution with respect to the filter.final classDepthwiseConv2dNativeBackpropInput<T extends TNumber>Computes the gradients of depthwise convolution with respect to the input.final classDilation2d<T extends TNumber>Computes the grayscale dilation of 4-Dinputand 3-Dfiltertensors.final classDilation2dBackpropFilter<T extends TNumber>Computes the gradient of morphological 2-D dilation with respect to the filter.final classDilation2dBackpropInput<T extends TNumber>Computes the gradient of morphological 2-D dilation with respect to the input.final classComputes the exponential linear function.final classComputes gradients for the exponential linear (Elu) operation.final classFractionalAvgPoolGrad<T extends TNumber>Computes gradient of the FractionalAvgPool function.final classFractionalMaxPoolGrad<T extends TNumber>Computes gradient of the FractionalMaxPool function.final classFusedPadConv2d<T extends TNumber>Performs a padding as a preprocess during a convolution.final classFusedResizeAndPadConv2d<T extends TNumber>Performs a resize and padding as a preprocess during a convolution.final classSays whether the targets are in the topKpredictions.final classComputes the gradient for the inverse ofxwrt its input.final classL2 Loss.final classComputes rectified linear:max(features, features * alpha).final classLocalResponseNormalization<T extends TNumber>Local Response Normalization.final classLocalResponseNormalizationGrad<T extends TNumber>Gradients for Local Response Normalization.final classLogSoftmax<T extends TNumber>Computes log softmax activations.final classPerforms max pooling on the input.final classPerforms 3D max pooling on the input.final classMaxPool3dGrad<U extends TNumber>Computes gradients of 3D max pooling function.final classMaxPool3dGradGrad<T extends TNumber>Computes second-order gradients of the maxpooling function.final classMaxPoolGrad<T extends TNumber>Computes gradients of the maxpooling function.final classMaxPoolGradGrad<T extends TNumber>Computes second-order gradients of the maxpooling function.final classMaxPoolGradGradWithArgmax<T extends TNumber>Computes second-order gradients of the maxpooling function.final classMaxPoolGradWithArgmax<T extends TNumber>Computes gradients of the maxpooling function.final classNthElement<T extends TNumber>Finds values of then-th order statistic for the last dimension.final classComputes rectified linear:max(features, 0).final classComputes rectified linear 6:min(max(features, 0), 6).final classComputes rectified linear 6 gradients for a Relu6 operation.final classComputes rectified linear gradients for a Relu operation.final classComputes scaled exponential linear:scale * alpha * (exp(features) - 1)if < 0,scale * featuresotherwise.final classComputes gradients for the scaled exponential linear (Selu) operation.final classComputes softmax activations.final classComputes softsign:features / (abs(features) + 1).final classSoftsignGrad<T extends TNumber>Computes softsign gradients for a softsign operation.final classSpaceToBatch<T extends TType>SpaceToBatch for 4-D tensors of type T.final classSpaceToDepth<T extends TType>SpaceToDepth for tensors of type T.final classUniformQuantizedConvolution<U extends TNumber>Perform quantized convolution of quantized Tensorlhsand quantized Tensorrhs. to make quantizedoutput.final classUniformQuantizedConvolutionHybrid<V extends TNumber>Perform hybrid quantized convolution of float Tensorlhsand quantized Tensorrhs.Fields in org.tensorflow.op.nn declared as OperandModifier and TypeFieldDescriptionMaxPoolGradGradWithArgmax.Inputs.argmaxThe indices of the maximum values chosen for each output ofmax_pool.MaxPoolGradWithArgmax.Inputs.argmaxThe indices of the maximum values chosen for each output ofmax_pool.BlockLSTM.Inputs.bThe bias vector.BlockLSTMGrad.Inputs.bThe bias vector.LSTMBlockCell.Inputs.bThe bias vector.LSTMBlockCellGrad.Inputs.bThe bias vector.BatchNormWithGlobalNormalizationGrad.Inputs.backprop4D backprop Tensor.GRUBlockCell.Inputs.bCThe bC inputGRUBlockCellGrad.Inputs.bCThe bC inputBatchNormWithGlobalNormalization.Inputs.betaA 1D beta Tensor with size matching the last dimension of t.QuantizedBatchNormWithGlobalNormalization.Inputs.betaA 1D beta Tensor with size matching the last dimension of t.QuantizedBatchNormWithGlobalNormalization.Inputs.betaMaxThe value represented by the highest quantized offset.QuantizedBatchNormWithGlobalNormalization.Inputs.betaMinThe value represented by the lowest quantized offset.BiasAdd.Inputs.bias1-D with size the last dimension ofvalue.QuantizedBiasAdd.Inputs.biasA 1D bias Tensor with size matching the last dimension of 'input'.QuantizedConv2DWithBias.Inputs.biasThe bias inputQuantizedConv2DWithBiasAndRelu.Inputs.biasThe bias inputQuantizedConv2DWithBiasAndReluAndRequantize.Inputs.biasThe bias inputQuantizedConv2DWithBiasAndRequantize.Inputs.biasThe bias inputQuantizedConv2DWithBiasSignedSumAndReluAndRequantize.Inputs.biasThe bias inputQuantizedConv2DWithBiasSumAndRelu.Inputs.biasThe bias inputQuantizedConv2DWithBiasSumAndReluAndRequantize.Inputs.biasThe bias inputQuantizedDepthwiseConv2DWithBias.Inputs.biasThe original bias tensor.QuantizedDepthwiseConv2DWithBiasAndRelu.Inputs.biasThe original bias tensor.QuantizedDepthwiseConv2DWithBiasAndReluAndRequantize.Inputs.biasThe original bias tensor.GRUBlockCell.Inputs.bRuThe bRu inputGRUBlockCellGrad.Inputs.bRuThe bRu inputGRUBlockCellGrad.Inputs.cThe c inputBlockLSTMGrad.Inputs.ciThe cell input over the whole time sequence.LSTMBlockCellGrad.Inputs.ciThe cell input.BlockLSTMGrad.Inputs.coThe cell after the tanh over the whole time sequence.LSTMBlockCellGrad.Inputs.coThe cell after the tanh.FractionalAvgPoolGrad.Inputs.colPoolingSequencecolumn pooling sequence, form pooling region with row_pooling sequence.FractionalMaxPoolGrad.Inputs.colPoolingSequencecolumn pooling sequence, form pooling region with row_pooling sequence.BlockLSTMGrad.Inputs.csThe cell state before the tanh over the whole time sequence.LSTMBlockCellGrad.Inputs.csThe cell state before the tanh.BlockLSTMGrad.Inputs.csGradThe current gradient of cs.LSTMBlockCellGrad.Inputs.csGradThe current gradient of cs.BlockLSTM.Inputs.csPrevValue of the initial cell state.BlockLSTMGrad.Inputs.csPrevValue of the initial cell state.LSTMBlockCell.Inputs.csPrevValue of the cell state at previous time step.LSTMBlockCellGrad.Inputs.csPrevThe previous cell state.GRUBlockCellGrad.Inputs.dHThe dH inputInvGrad.Inputs.dyThe dy inputBlockLSTMGrad.Inputs.fThe forget gate over the whole time sequence.LSTMBlockCellGrad.Inputs.fThe forget gate.Elu.Inputs.featuresThe features inputLeakyRelu.Inputs.featuresThe features inputQuantizedRelu.Inputs.featuresThe features inputQuantizedRelu6.Inputs.featuresThe features inputQuantizedReluX.Inputs.featuresThe features inputRelu.Inputs.featuresThe features inputRelu6.Inputs.featuresThe features inputRelu6Grad.Inputs.featuresThe features passed as input to the corresponding Relu6 operation, or its output; using either one produces the same result.ReluGrad.Inputs.featuresThe features passed as input to the corresponding Relu operation, OR the outputs of that operation (both work equivalently).Selu.Inputs.featuresThe features inputSoftmaxCrossEntropyWithLogits.Inputs.featuresbatch_size x num_classes matrixSoftsign.Inputs.featuresThe features inputSoftsignGrad.Inputs.featuresThe features passed as input to the corresponding softsign operation.SparseSoftmaxCrossEntropyWithLogits.Inputs.featuresbatch_size x num_classes matrixConv.Inputs.filterAn(N+2)-DTensor with the same type asinputand shapespatial_filter_shape + [in_channels, out_channels], where spatial_filter_shape is N-dimensional withN=2orN=3.Conv2d.Inputs.filterA 4-D tensor of shape[filter_height, filter_width, in_channels, out_channels]Conv2dBackpropFilterV2.Inputs.filter4-D with shape[filter_height, filter_width, in_channels, out_channels].Conv2dBackpropInput.Inputs.filter4-D with shape[filter_height, filter_width, in_channels, out_channels].Conv2dBackpropInputV2.Inputs.filter4-D with shape[filter_height, filter_width, in_channels, out_channels].Conv3d.Inputs.filterShape[filter_depth, filter_height, filter_width, in_channels, out_channels].Conv3dBackpropInput.Inputs.filterShape[depth, rows, cols, in_channels, out_channels].DepthwiseConv2dNative.Inputs.filterThe filter inputDepthwiseConv2dNativeBackpropInput.Inputs.filter4-D with shape[filter_height, filter_width, in_channels, depthwise_multiplier].Dilation2d.Inputs.filter3-D with shape[filter_height, filter_width, depth].Dilation2dBackpropFilter.Inputs.filter3-D with shape[filter_height, filter_width, depth].Dilation2dBackpropInput.Inputs.filter3-D with shape[filter_height, filter_width, depth].FusedPadConv2d.Inputs.filter4-D with shape[filter_height, filter_width, in_channels, out_channels].FusedResizeAndPadConv2d.Inputs.filter4-D with shape[filter_height, filter_width, in_channels, out_channels].QuantizedConv2d.Inputs.filterfilter's input_depth dimension must match input's depth dimensions.QuantizedConv2DAndRelu.Inputs.filterThe filter inputQuantizedConv2DAndReluAndRequantize.Inputs.filterThe filter inputQuantizedConv2DAndRequantize.Inputs.filterThe filter inputQuantizedConv2DPerChannel.Inputs.filterThe original filter tensor.QuantizedConv2DWithBias.Inputs.filterThe filter inputQuantizedConv2DWithBiasAndRelu.Inputs.filterThe filter inputQuantizedConv2DWithBiasAndReluAndRequantize.Inputs.filterThe filter inputQuantizedConv2DWithBiasAndRequantize.Inputs.filterThe filter inputQuantizedConv2DWithBiasSignedSumAndReluAndRequantize.Inputs.filterThe filter inputQuantizedConv2DWithBiasSumAndRelu.Inputs.filterThe filter inputQuantizedConv2DWithBiasSumAndReluAndRequantize.Inputs.filterThe filter inputQuantizedDepthwiseConv2D.Inputs.filterThe original filter tensor.QuantizedDepthwiseConv2DWithBias.Inputs.filterThe original filter tensor.QuantizedDepthwiseConv2DWithBiasAndRelu.Inputs.filterThe original filter tensor.QuantizedDepthwiseConv2DWithBiasAndReluAndRequantize.Inputs.filterThe original filter tensor.Conv2dBackpropFilter.Inputs.filterSizesAn integer vector representing the tensor shape offilter, wherefilteris a 4-D[filter_height, filter_width, in_channels, out_channels]tensor.Conv3dBackpropFilter.Inputs.filterSizesAn integer vector representing the tensor shape offilter, wherefilteris a 5-D[filter_depth, filter_height, filter_width, in_channels, out_channels]tensor.DepthwiseConv2dNativeBackpropFilter.Inputs.filterSizesAn integer vector representing the tensor shape offilter, wherefilteris a 4-D[filter_height, filter_width, in_channels, depthwise_multiplier]tensor.BatchNormWithGlobalNormalization.Inputs.gammaA 1D gamma Tensor with size matching the last dimension of t.BatchNormWithGlobalNormalizationGrad.Inputs.gammaA 1D gamma Tensor with size matching the last dimension of t.QuantizedBatchNormWithGlobalNormalization.Inputs.gammaA 1D gamma Tensor with size matching the last dimension of t.QuantizedBatchNormWithGlobalNormalization.Inputs.gammaMaxThe value represented by the highest quantized gamma.QuantizedBatchNormWithGlobalNormalization.Inputs.gammaMinThe value represented by the lowest quantized gamma.AvgPool3dGrad.Inputs.gradOutput backprop of shape[batch, depth, rows, cols, channels].AvgPoolGrad.Inputs.grad4-D with shape[batch, height, width, channels].MaxPool3dGrad.Inputs.gradOutput backprop of shape[batch, depth, rows, cols, channels].MaxPool3dGradGrad.Inputs.gradOutput backprop of shape[batch, depth, rows, cols, channels].MaxPoolGrad.Inputs.grad4-D.MaxPoolGradGrad.Inputs.grad4-D.MaxPoolGradGradWithArgmax.Inputs.grad4-D with shape[batch, height, width, channels].MaxPoolGradWithArgmax.Inputs.grad4-D with shape[batch, height, width, channels].EluGrad.Inputs.gradientsThe backpropagated gradients to the corresponding Elu operation.Relu6Grad.Inputs.gradientsThe backpropagated gradients to the corresponding Relu6 operation.ReluGrad.Inputs.gradientsThe backpropagated gradients to the corresponding Relu operation.SeluGrad.Inputs.gradientsThe backpropagated gradients to the corresponding Selu operation.SoftsignGrad.Inputs.gradientsThe backpropagated gradients to the corresponding softsign operation.BlockLSTMGrad.Inputs.hThe output h vector over the whole time sequence.BlockLSTMGrad.Inputs.hGradThe gradient of h vector.LSTMBlockCellGrad.Inputs.hGradThe gradient of h vector.CudnnRNNBackprop.Inputs.hostReservedThe hostReserved inputBlockLSTM.Inputs.hPrevInitial output of cell (to be used for peephole).BlockLSTMGrad.Inputs.hPrevInitial output of cell (to be used for peephole).GRUBlockCell.Inputs.hPrevThe hPrev inputGRUBlockCellGrad.Inputs.hPrevThe hPrev inputLSTMBlockCell.Inputs.hPrevOutput of the previous cell at previous time step.LSTMBlockCellGrad.Inputs.hPrevThe previous h state.BlockLSTMGrad.Inputs.iThe input gate over the whole time sequence.LSTMBlockCellGrad.Inputs.iThe input gate.AvgPool3d.Inputs.inputShape[batch, depth, rows, cols, channels]tensor to pool over.Conv.Inputs.inputTensor of type T and shapebatch_shape + spatial_shape + [in_channels]in the case thatchannels_last_format = trueor shapebatch_shape + [in_channels] + spatial_shapeifchannels_last_format = false. spatial_shape is N-dimensional withN=2orN=3.Conv2d.Inputs.inputA 4-D tensor.Conv2dBackpropFilter.Inputs.input4-D with shape[batch, in_height, in_width, in_channels].Conv2dBackpropFilterV2.Inputs.input4-D with shape[batch, in_height, in_width, in_channels].Conv2dBackpropInputV2.Inputs.input4-D with shape[batch, in_height, in_width, in_channels].Conv3d.Inputs.inputShape[batch, in_depth, in_height, in_width, in_channels].Conv3dBackpropFilter.Inputs.inputShape[batch, depth, rows, cols, in_channels].CudnnRNN.Inputs.inputThe input inputCudnnRNNBackprop.Inputs.inputThe input inputDepthToSpace.Inputs.inputThe input inputDepthwiseConv2dNative.Inputs.inputThe input inputDepthwiseConv2dNativeBackpropFilter.Inputs.input4-D with shape based ondata_format.Dilation2d.Inputs.input4-D with shape[batch, in_height, in_width, depth].Dilation2dBackpropFilter.Inputs.input4-D with shape[batch, in_height, in_width, depth].Dilation2dBackpropInput.Inputs.input4-D with shape[batch, in_height, in_width, depth].FusedPadConv2d.Inputs.input4-D with shape[batch, in_height, in_width, in_channels].FusedResizeAndPadConv2d.Inputs.input4-D with shape[batch, in_height, in_width, in_channels].IsotonicRegression.Inputs.inputA (batch_size, dim)-tensor holding a batch of inputs.LocalResponseNormalization.Inputs.input4-D.MaxPool.Inputs.input4-D input to pool over.MaxPool3d.Inputs.inputShape[batch, depth, rows, cols, channels]tensor to pool over.MaxPoolGradGradWithArgmax.Inputs.inputThe original input.MaxPoolGradWithArgmax.Inputs.inputThe original input.MaxPoolWithArgmax.Inputs.input4-D with shape[batch, height, width, channels].NthElement.Inputs.input1-D or higher with last dimension at leastn+1.QuantizedAvgPool.Inputs.input4-D with shape[batch, height, width, channels].QuantizedBiasAdd.Inputs.inputThe input inputQuantizedConv2d.Inputs.inputThe input inputQuantizedConv2DAndRelu.Inputs.inputThe input inputQuantizedConv2DAndReluAndRequantize.Inputs.inputThe input inputQuantizedConv2DAndRequantize.Inputs.inputThe input inputQuantizedConv2DPerChannel.Inputs.inputThe original input tensor.QuantizedConv2DWithBias.Inputs.inputThe input inputQuantizedConv2DWithBiasAndRelu.Inputs.inputThe input inputQuantizedConv2DWithBiasAndReluAndRequantize.Inputs.inputThe input inputQuantizedConv2DWithBiasAndRequantize.Inputs.inputThe input inputQuantizedConv2DWithBiasSignedSumAndReluAndRequantize.Inputs.inputThe input inputQuantizedConv2DWithBiasSumAndRelu.Inputs.inputThe input inputQuantizedConv2DWithBiasSumAndReluAndRequantize.Inputs.inputThe input inputQuantizedDepthwiseConv2D.Inputs.inputThe original input tensor.QuantizedDepthwiseConv2DWithBias.Inputs.inputThe original input tensor.QuantizedDepthwiseConv2DWithBiasAndRelu.Inputs.inputThe original input tensor.QuantizedDepthwiseConv2DWithBiasAndReluAndRequantize.Inputs.inputThe original input tensor.QuantizedMaxPool.Inputs.inputThe 4D (batch x rows x cols x depth) Tensor to MaxReduce over.SpaceToBatch.Inputs.input4-D with shape[batch, height, width, depth].SpaceToDepth.Inputs.inputThe input inputTopK.Inputs.input1-D or higher with last dimension at leastk.CudnnRNN.Inputs.inputCThe inputC inputCudnnRNNBackprop.Inputs.inputCThe inputC inputLocalResponseNormalizationGrad.Inputs.inputGrads4-D with shape[batch, height, width, channels].CudnnRNN.Inputs.inputHThe inputH inputCudnnRNNBackprop.Inputs.inputHThe inputH inputLocalResponseNormalizationGrad.Inputs.inputImage4-D with shape[batch, height, width, channels].CtcBeamSearchDecoder.Inputs.inputs3-D, shape:(max_time x batch_size x num_classes), the logits.CtcGreedyDecoder.Inputs.inputs3-D, shape:(max_time x batch_size x num_classes), the logits.CtcLoss.Inputs.inputs3-D, shape:(max_time x batch_size x num_classes), the logits.CTCLossV2.Inputs.inputs3-D, shape:(max_time x batch_size x num_classes), the logits.CudnnRNNCanonicalToParams.Inputs.inputSizeThe inputSize inputCudnnRnnParamsSize.Inputs.inputSizeThe inputSize inputCudnnRNNParamsToCanonical.Inputs.inputSizeThe inputSize inputConv2dBackpropInput.Inputs.inputSizesAn integer vector representing the shape ofinput, whereinputis a 4-D[batch, height, width, channels]tensor.Conv3dBackpropInput.Inputs.inputSizesAn integer vector representing the tensor shape ofinput, whereinputis a 5-D[batch, depth, rows, cols, in_channels]tensor.DepthwiseConv2dNativeBackpropInput.Inputs.inputSizesAn integer vector representing the shape ofinput, based ondata_format.InTopK.Inputs.kNumber of top elements to look at for computing precision.TopK.Inputs.k0-D.MaxPool.Inputs.ksizeThe size of the window for each dimension of the input tensor.MaxPoolGrad.Inputs.ksizeThe size of the window for each dimension of the input tensor.MaxPoolGradGrad.Inputs.ksizeThe size of the window for each dimension of the input tensor.SoftmaxCrossEntropyWithLogits.Inputs.labelsbatch_size x num_classes matrix The caller must ensure that each batch of labels represents a valid probability distribution.SparseSoftmaxCrossEntropyWithLogits.Inputs.labelsbatch_size vector with values in [0, num_classes).CtcLoss.Inputs.labelsIndicesThe indices of aSparseTensor<int32, 2>.CTCLossV2.Inputs.labelsIndicesThe indices of aSparseTensor<int32, 2>.CtcLoss.Inputs.labelsValuesThe values (labels) associated with the given batch and time.CTCLossV2.Inputs.labelsValuesThe values (labels) associated with the given batch and time.UniformQuantizedConvolution.Inputs.lhsMust be a quantized tensor, rank >= 3.UniformQuantizedConvolutionHybrid.Inputs.lhsMust be a non-quantized Tensor ofTlhs, rank >= 3.UniformQuantizedConvolution.Inputs.lhsScalesThe float value(s) used as scale factors when quantizing the original data thatlhsrepresents.UniformQuantizedConvolution.Inputs.lhsZeroPointsThe int32 value(s) used as zero points when quantizing original data thatlhsrepresents.LogSoftmax.Inputs.logits2-D with shape[batch_size, num_classes].Softmax.Inputs.logits2-D with shape[batch_size, num_classes].BatchNormWithGlobalNormalization.Inputs.mA 1D mean Tensor with size matching the last dimension of t.BatchNormWithGlobalNormalizationGrad.Inputs.mA 1D mean Tensor with size matching the last dimension of t.QuantizedBatchNormWithGlobalNormalization.Inputs.mA 1D mean Tensor with size matching the last dimension of t.QuantizedBiasAdd.Inputs.maxBiasThe float value that the highest quantized bias value represents.QuantizedRelu.Inputs.maxFeaturesThe float value that the highest quantized value represents.QuantizedRelu6.Inputs.maxFeaturesThe float value that the highest quantized value represents.QuantizedReluX.Inputs.maxFeaturesThe float value that the highest quantized value represents.QuantizedConv2d.Inputs.maxFilterThe float value that the highest quantized filter value represents.QuantizedConv2DAndRelu.Inputs.maxFilterThe maxFilter inputQuantizedConv2DAndReluAndRequantize.Inputs.maxFilterThe maxFilter inputQuantizedConv2DAndRequantize.Inputs.maxFilterThe maxFilter inputQuantizedConv2DPerChannel.Inputs.maxFilterThe maximum value of the filter tensor.QuantizedConv2DWithBias.Inputs.maxFilterThe maxFilter inputQuantizedConv2DWithBiasAndRelu.Inputs.maxFilterThe maxFilter inputQuantizedConv2DWithBiasAndReluAndRequantize.Inputs.maxFilterThe maxFilter inputQuantizedConv2DWithBiasAndRequantize.Inputs.maxFilterThe maxFilter inputQuantizedConv2DWithBiasSignedSumAndReluAndRequantize.Inputs.maxFilterThe maxFilter inputQuantizedConv2DWithBiasSumAndRelu.Inputs.maxFilterThe maxFilter inputQuantizedConv2DWithBiasSumAndReluAndRequantize.Inputs.maxFilterThe maxFilter inputQuantizedDepthwiseConv2D.Inputs.maxFilterThe float value that the maximum quantized filter value represents.QuantizedDepthwiseConv2DWithBias.Inputs.maxFilterThe float value that the maximum quantized filter value represents.QuantizedDepthwiseConv2DWithBiasAndRelu.Inputs.maxFilterThe float value that the maximum quantized filter value represents.QuantizedDepthwiseConv2DWithBiasAndReluAndRequantize.Inputs.maxFilterThe float value that the maximum quantized filter value represents.QuantizedConv2DAndReluAndRequantize.Inputs.maxFreezedOutputThe maxFreezedOutput inputQuantizedConv2DAndRequantize.Inputs.maxFreezedOutputThe maxFreezedOutput inputQuantizedConv2DWithBiasAndReluAndRequantize.Inputs.maxFreezedOutputThe maxFreezedOutput inputQuantizedConv2DWithBiasAndRequantize.Inputs.maxFreezedOutputThe maxFreezedOutput inputQuantizedConv2DWithBiasSignedSumAndReluAndRequantize.Inputs.maxFreezedOutputThe maxFreezedOutput inputQuantizedConv2DWithBiasSumAndReluAndRequantize.Inputs.maxFreezedOutputThe maxFreezedOutput inputQuantizedDepthwiseConv2DWithBiasAndReluAndRequantize.Inputs.maxFreezedOutputThe maximum float value of the output tensor.QuantizedAvgPool.Inputs.maxInputThe float value that the highest quantized input value represents.QuantizedBiasAdd.Inputs.maxInputThe float value that the highest quantized input value represents.QuantizedConv2d.Inputs.maxInputThe float value that the highest quantized input value represents.QuantizedConv2DAndRelu.Inputs.maxInputThe maxInput inputQuantizedConv2DAndReluAndRequantize.Inputs.maxInputThe maxInput inputQuantizedConv2DAndRequantize.Inputs.maxInputThe maxInput inputQuantizedConv2DPerChannel.Inputs.maxInputThe maximum value of the input tensor.QuantizedConv2DWithBias.Inputs.maxInputThe maxInput inputQuantizedConv2DWithBiasAndRelu.Inputs.maxInputThe maxInput inputQuantizedConv2DWithBiasAndReluAndRequantize.Inputs.maxInputThe maxInput inputQuantizedConv2DWithBiasAndRequantize.Inputs.maxInputThe maxInput inputQuantizedConv2DWithBiasSignedSumAndReluAndRequantize.Inputs.maxInputThe maxInput inputQuantizedConv2DWithBiasSumAndRelu.Inputs.maxInputThe maxInput inputQuantizedConv2DWithBiasSumAndReluAndRequantize.Inputs.maxInputThe maxInput inputQuantizedDepthwiseConv2D.Inputs.maxInputThe float value that the maximum quantized input value represents.QuantizedDepthwiseConv2DWithBias.Inputs.maxInputThe float value that the maximum quantized input value represents.QuantizedDepthwiseConv2DWithBiasAndRelu.Inputs.maxInputThe float value that the maximum quantized input value represents.QuantizedDepthwiseConv2DWithBiasAndReluAndRequantize.Inputs.maxInputThe float value that the maximum quantized input value represents.QuantizedMaxPool.Inputs.maxInputThe float value that the highest quantized input value represents.QuantizedConv2DWithBiasSignedSumAndReluAndRequantize.Inputs.maxSummandThe maxSummand inputQuantizedConv2DWithBiasSumAndReluAndRequantize.Inputs.maxSummandThe maxSummand inputQuantizedReluX.Inputs.maxValueThe maxValue inputFusedBatchNorm.Inputs.meanA 1D Tensor for population mean.QuantizedBiasAdd.Inputs.minBiasThe float value that the lowest quantized bias value represents.QuantizedRelu.Inputs.minFeaturesThe float value that the lowest quantized value represents.QuantizedRelu6.Inputs.minFeaturesThe float value that the lowest quantized value represents.QuantizedReluX.Inputs.minFeaturesThe float value that the lowest quantized value represents.QuantizedConv2d.Inputs.minFilterThe float value that the lowest quantized filter value represents.QuantizedConv2DAndRelu.Inputs.minFilterThe minFilter inputQuantizedConv2DAndReluAndRequantize.Inputs.minFilterThe minFilter inputQuantizedConv2DAndRequantize.Inputs.minFilterThe minFilter inputQuantizedConv2DPerChannel.Inputs.minFilterThe minimum value of the filter tensor.QuantizedConv2DWithBias.Inputs.minFilterThe minFilter inputQuantizedConv2DWithBiasAndRelu.Inputs.minFilterThe minFilter inputQuantizedConv2DWithBiasAndReluAndRequantize.Inputs.minFilterThe minFilter inputQuantizedConv2DWithBiasAndRequantize.Inputs.minFilterThe minFilter inputQuantizedConv2DWithBiasSignedSumAndReluAndRequantize.Inputs.minFilterThe minFilter inputQuantizedConv2DWithBiasSumAndRelu.Inputs.minFilterThe minFilter inputQuantizedConv2DWithBiasSumAndReluAndRequantize.Inputs.minFilterThe minFilter inputQuantizedDepthwiseConv2D.Inputs.minFilterThe float value that the minimum quantized filter value represents.QuantizedDepthwiseConv2DWithBias.Inputs.minFilterThe float value that the minimum quantized filter value represents.QuantizedDepthwiseConv2DWithBiasAndRelu.Inputs.minFilterThe float value that the minimum quantized filter value represents.QuantizedDepthwiseConv2DWithBiasAndReluAndRequantize.Inputs.minFilterThe float value that the minimum quantized filter value represents.QuantizedConv2DAndReluAndRequantize.Inputs.minFreezedOutputThe minFreezedOutput inputQuantizedConv2DAndRequantize.Inputs.minFreezedOutputThe minFreezedOutput inputQuantizedConv2DWithBiasAndReluAndRequantize.Inputs.minFreezedOutputThe minFreezedOutput inputQuantizedConv2DWithBiasAndRequantize.Inputs.minFreezedOutputThe minFreezedOutput inputQuantizedConv2DWithBiasSignedSumAndReluAndRequantize.Inputs.minFreezedOutputThe minFreezedOutput inputQuantizedConv2DWithBiasSumAndReluAndRequantize.Inputs.minFreezedOutputThe minFreezedOutput inputQuantizedDepthwiseConv2DWithBiasAndReluAndRequantize.Inputs.minFreezedOutputThe minimum float value of the output tensor.QuantizedAvgPool.Inputs.minInputThe float value that the lowest quantized input value represents.QuantizedBiasAdd.Inputs.minInputThe float value that the lowest quantized input value represents.QuantizedConv2d.Inputs.minInputThe float value that the lowest quantized input value represents.QuantizedConv2DAndRelu.Inputs.minInputThe minInput inputQuantizedConv2DAndReluAndRequantize.Inputs.minInputThe minInput inputQuantizedConv2DAndRequantize.Inputs.minInputThe minInput inputQuantizedConv2DPerChannel.Inputs.minInputThe minimum value of the input tensorQuantizedConv2DWithBias.Inputs.minInputThe minInput inputQuantizedConv2DWithBiasAndRelu.Inputs.minInputThe minInput inputQuantizedConv2DWithBiasAndReluAndRequantize.Inputs.minInputThe minInput inputQuantizedConv2DWithBiasAndRequantize.Inputs.minInputThe minInput inputQuantizedConv2DWithBiasSignedSumAndReluAndRequantize.Inputs.minInputThe minInput inputQuantizedConv2DWithBiasSumAndRelu.Inputs.minInputThe minInput inputQuantizedConv2DWithBiasSumAndReluAndRequantize.Inputs.minInputThe minInput inputQuantizedDepthwiseConv2D.Inputs.minInputThe float value that the minimum quantized input value represents.QuantizedDepthwiseConv2DWithBias.Inputs.minInputThe float value that the minimum quantized input value represents.QuantizedDepthwiseConv2DWithBiasAndRelu.Inputs.minInputThe float value that the minimum quantized input value represents.QuantizedDepthwiseConv2DWithBiasAndReluAndRequantize.Inputs.minInputThe float value that the minimum quantized input value represents.QuantizedMaxPool.Inputs.minInputThe float value that the lowest quantized input value represents.QuantizedConv2DWithBiasSignedSumAndReluAndRequantize.Inputs.minSummandThe minSummand inputQuantizedConv2DWithBiasSumAndReluAndRequantize.Inputs.minSummandThe minSummand inputQuantizedBatchNormWithGlobalNormalization.Inputs.mMaxThe value represented by the highest quantized mean.QuantizedBatchNormWithGlobalNormalization.Inputs.mMinThe value represented by the lowest quantized mean.NthElement.Inputs.n0-D.CudnnRNNCanonicalToParams.Inputs.numLayersThe numLayers inputCudnnRnnParamsSize.Inputs.numLayersThe numLayers inputCudnnRNNParamsToCanonical.Inputs.numLayersThe numLayers inputCudnnRNNCanonicalToParams.Inputs.numUnitsThe numUnits inputCudnnRnnParamsSize.Inputs.numUnitsThe numUnits inputCudnnRNNParamsToCanonical.Inputs.numUnitsThe numUnits inputBlockLSTMGrad.Inputs.oThe output gate over the whole time sequence.LSTMBlockCellGrad.Inputs.oThe output gate.FusedBatchNorm.Inputs.offsetA 1D Tensor for offset, to shift to the normalized x.FractionalMaxPoolGrad.Inputs.origInputOriginal input forfractional_max_poolMaxPool3dGrad.Inputs.origInputThe original input tensor.MaxPool3dGradGrad.Inputs.origInputThe original input tensor.MaxPoolGrad.Inputs.origInputThe original input tensor.MaxPoolGradGrad.Inputs.origInputThe original input tensor.AvgPool3dGrad.Inputs.origInputShapeThe original input dimensions.AvgPoolGrad.Inputs.origInputShape1-D.FractionalAvgPoolGrad.Inputs.origInputTensorShapeOriginal input tensor shape forfractional_avg_poolFractionalMaxPoolGrad.Inputs.origOutputOriginal output forfractional_max_poolMaxPool3dGrad.Inputs.origOutputThe original output tensor.MaxPool3dGradGrad.Inputs.origOutputThe original output tensor.MaxPoolGrad.Inputs.origOutputThe original output tensor.MaxPoolGradGrad.Inputs.origOutputThe original output tensor.BiasAddGrad.Inputs.outBackpropAny number of dimensions.Conv2dBackpropFilter.Inputs.outBackprop4-D with shape[batch, out_height, out_width, out_channels].Conv2dBackpropFilterV2.Inputs.outBackprop4-D with shape[batch, out_height, out_width, out_channels].Conv2dBackpropInput.Inputs.outBackprop4-D with shape[batch, out_height, out_width, out_channels].Conv2dBackpropInputV2.Inputs.outBackprop4-D with shape[batch, out_height, out_width, out_channels].Conv3dBackpropFilter.Inputs.outBackpropBackprop signal of shape[batch, out_depth, out_rows, out_cols, out_channels].Conv3dBackpropInput.Inputs.outBackpropBackprop signal of shape[batch, out_depth, out_rows, out_cols, out_channels].DepthwiseConv2dNativeBackpropFilter.Inputs.outBackprop4-D with shape based ondata_format.DepthwiseConv2dNativeBackpropInput.Inputs.outBackprop4-D with shape based ondata_format.Dilation2dBackpropFilter.Inputs.outBackprop4-D with shape[batch, out_height, out_width, depth].Dilation2dBackpropInput.Inputs.outBackprop4-D with shape[batch, out_height, out_width, depth].FractionalAvgPoolGrad.Inputs.outBackprop4-D with shape[batch, height, width, channels].FractionalMaxPoolGrad.Inputs.outBackprop4-D with shape[batch, height, width, channels].CudnnRNNBackprop.Inputs.outputThe output inputCudnnRNNBackprop.Inputs.outputBackpropThe outputBackprop inputCudnnRNNBackprop.Inputs.outputCThe outputC inputCudnnRNNBackprop.Inputs.outputCBackpropThe outputCBackprop inputCudnnRNNBackprop.Inputs.outputHThe outputH inputCudnnRNNBackprop.Inputs.outputHBackpropThe outputHBackprop inputLocalResponseNormalizationGrad.Inputs.outputImage4-D with shape[batch, height, width, channels].EluGrad.Inputs.outputsThe outputs of the corresponding Elu operation.SeluGrad.Inputs.outputsThe outputs of the corresponding Selu operation.UniformQuantizedConvolution.Inputs.outputScalesThe float value(s) to use as scale factors when quantizing original data thatoutputrepresents.UniformQuantizedConvolution.Inputs.outputZeroPointsThe int32 value(s) used as zero points when quantizing original data that output represents.FusedPadConv2d.Inputs.paddingsA two-column matrix specifying the padding sizes.FusedResizeAndPadConv2d.Inputs.paddingsA two-column matrix specifying the padding sizes.SpaceToBatch.Inputs.paddings2-D tensor of non-negative integers with shape[2, 2].CudnnRNN.Inputs.paramsThe params inputCudnnRNNBackprop.Inputs.paramsThe params inputCudnnRNNParamsToCanonical.Inputs.paramsThe params inputInTopK.Inputs.predictionsAbatch_sizexclassestensor.GRUBlockCellGrad.Inputs.rThe r inputCudnnRNNBackprop.Inputs.reserveSpaceThe reserveSpace inputFusedBatchNormGrad.Inputs.reserveSpace1When is_training is True, a 1D Tensor for the computed batch mean to be reused in gradient computation.FusedBatchNormGrad.Inputs.reserveSpace2When is_training is True, a 1D Tensor for the computed batch variance (inverted variance in the cuDNN case) to be reused in gradient computation.FusedBatchNormGrad.Inputs.reserveSpace3When is_training is True, a 1D Tensor for some intermediate results to be reused in gradient computation.UniformQuantizedConvolution.Inputs.rhsMust be a quantized tensor, same rank aslhs.UniformQuantizedConvolutionHybrid.Inputs.rhsMust be a quantized Tensor ofTrhs, same rank aslhs.UniformQuantizedConvolution.Inputs.rhsScalesThe float value(s) used as scale factors when quantizing the original data thatrhsrepresents.UniformQuantizedConvolutionHybrid.Inputs.rhsScalesThe float value(s) used as scale factors when quantizing the original data thatrhsrepresents.UniformQuantizedConvolution.Inputs.rhsZeroPointsThe int32 value(s) used as zero points when quantizing original data thatrhsrepresents.UniformQuantizedConvolutionHybrid.Inputs.rhsZeroPointsThe int32 value(s) used as zero_point when quantizing original data thatrhsrepresents.FractionalAvgPoolGrad.Inputs.rowPoolingSequencerow pooling sequence, form pooling region with col_pooling_sequence.FractionalMaxPoolGrad.Inputs.rowPoolingSequencerow pooling sequence, form pooling region with col_pooling_sequence.ComputeAccidentalHits.Inputs.sampledCandidatesThe sampled_candidates output of CandidateSampler.FusedBatchNorm.Inputs.scaleA 1D Tensor for scaling factor, to scale the normalized x.FusedBatchNormGrad.Inputs.scaleA 1D Tensor for scaling factor, to scale the normalized x.BlockLSTM.Inputs.seqLenMaxMaximum time length actually used by this input.BlockLSTMGrad.Inputs.seqLenMaxMaximum time length actually used by this input.CtcBeamSearchDecoder.Inputs.sequenceLengthA vector containing sequence lengths, size(batch).CtcGreedyDecoder.Inputs.sequenceLengthA vector containing sequence lengths, size(batch_size).CtcLoss.Inputs.sequenceLengthA vector containing sequence lengths (batch).CTCLossV2.Inputs.sequenceLengthA vector containing sequence lengths (batch).CudnnRNN.Inputs.sequenceLengthsThe sequenceLengths inputCudnnRNNBackprop.Inputs.sequenceLengthsThe sequenceLengths inputFusedResizeAndPadConv2d.Inputs.sizeOutputA 1-D int32 Tensor of 2 elements:new_height, new_width.MaxPool.Inputs.stridesThe stride of the sliding window for each dimension of the input tensor.MaxPoolGrad.Inputs.stridesThe stride of the sliding window for each dimension of the input tensor.MaxPoolGradGrad.Inputs.stridesThe stride of the sliding window for each dimension of the input tensor.QuantizedConv2DWithBiasSignedSumAndReluAndRequantize.Inputs.summandThe summand inputQuantizedConv2DWithBiasSumAndRelu.Inputs.summandThe summand inputQuantizedConv2DWithBiasSumAndReluAndRequantize.Inputs.summandThe summand inputBatchNormWithGlobalNormalization.Inputs.tA 4D input Tensor.BatchNormWithGlobalNormalizationGrad.Inputs.tA 4D input Tensor.L2Loss.Inputs.tTypically 2-D, but may have any dimensions.QuantizedBatchNormWithGlobalNormalization.Inputs.tA 4D input Tensor.InTopK.Inputs.targetsAbatch_sizevector of class ids.QuantizedBatchNormWithGlobalNormalization.Inputs.tMaxThe value represented by the highest quantized input.QuantizedBatchNormWithGlobalNormalization.Inputs.tMinThe value represented by the lowest quantized input.ComputeAccidentalHits.Inputs.trueClassesThe true_classes output of UnpackSparseLabels.FixedUnigramCandidateSampler.Inputs.trueClassesA batch_size * num_true matrix, in which each row contains the IDs of the num_true target_classes in the corresponding original label.LearnedUnigramCandidateSampler.Inputs.trueClassesA batch_size * num_true matrix, in which each row contains the IDs of the num_true target_classes in the corresponding original label.GRUBlockCellGrad.Inputs.uThe u inputBatchNormWithGlobalNormalization.Inputs.vA 1D variance Tensor with size matching the last dimension of t.BatchNormWithGlobalNormalizationGrad.Inputs.vA 1D variance Tensor with size matching the last dimension of t.QuantizedBatchNormWithGlobalNormalization.Inputs.vA 1D variance Tensor with size matching the last dimension of t.AvgPool.Inputs.value4-D with shape[batch, height, width, channels].BiasAdd.Inputs.valueAny number of dimensions.FractionalAvgPool.Inputs.value4-D with shape[batch, height, width, channels].FractionalMaxPool.Inputs.value4-D with shape[batch, height, width, channels].FusedBatchNorm.Inputs.varianceA 1D Tensor for population variance.QuantizedBatchNormWithGlobalNormalization.Inputs.vMaxThe value represented by the highest quantized variance.QuantizedBatchNormWithGlobalNormalization.Inputs.vMinThe value represented by the lowest quantized variance.BlockLSTM.Inputs.wThe weight matrix.BlockLSTMGrad.Inputs.wThe weight matrix.LSTMBlockCell.Inputs.wThe weight matrix.LSTMBlockCellGrad.Inputs.wThe weight matrix.GRUBlockCell.Inputs.wCThe wC inputGRUBlockCellGrad.Inputs.wCThe wC inputBlockLSTM.Inputs.wcfThe weight matrix for forget gate peephole connection.BlockLSTMGrad.Inputs.wcfThe weight matrix for forget gate peephole connection.LSTMBlockCell.Inputs.wcfThe weight matrix for forget gate peephole connection.LSTMBlockCellGrad.Inputs.wcfThe weight matrix for forget gate peephole connection.BlockLSTM.Inputs.wciThe weight matrix for input gate peephole connection.BlockLSTMGrad.Inputs.wciThe weight matrix for input gate peephole connection.LSTMBlockCell.Inputs.wciThe weight matrix for input gate peephole connection.LSTMBlockCellGrad.Inputs.wciThe weight matrix for input gate peephole connection.BlockLSTM.Inputs.wcoThe weight matrix for output gate peephole connection.BlockLSTMGrad.Inputs.wcoThe weight matrix for output gate peephole connection.LSTMBlockCell.Inputs.wcoThe weight matrix for output gate peephole connection.LSTMBlockCellGrad.Inputs.wcoThe weight matrix for output gate peephole connection.GRUBlockCell.Inputs.wRuThe wRu inputGRUBlockCellGrad.Inputs.wRuThe wRu inputBlockLSTM.Inputs.xThe sequence input to the LSTM, shape (timelen, batch_size, num_inputs).BlockLSTMGrad.Inputs.xThe sequence input to the LSTM, shape (timelen, batch_size, num_inputs).DataFormatDimMap.Inputs.xA Tensor with each element as a dimension index in source data format.DataFormatVecPermute.Inputs.xTensor of rank 1 or 2 in source data format.FusedBatchNorm.Inputs.xA 4D Tensor for input data.FusedBatchNormGrad.Inputs.xA 4D Tensor for input data.GRUBlockCell.Inputs.xThe x inputGRUBlockCellGrad.Inputs.xThe x inputLSTMBlockCell.Inputs.xThe input to the LSTM cell, shape (batch_size, num_inputs).LSTMBlockCellGrad.Inputs.xThe input to the LSTM cell, shape (batch_size, num_inputs).QuantizedInstanceNorm.Inputs.xA 4D input Tensor.QuantizedInstanceNorm.Inputs.xMaxThe value represented by the highest quantized input.QuantizedInstanceNorm.Inputs.xMinThe value represented by the lowest quantized input.InvGrad.Inputs.yThe y inputFusedBatchNormGrad.Inputs.yBackpropA 4D Tensor for the gradient with respect to y.Fields in org.tensorflow.op.nn with type parameters of type OperandModifier and TypeFieldDescriptionCudnnRNNCanonicalToParams.Inputs.biasesThe biases inputCudnnRNNCanonicalToParams.Inputs.weightsThe weights inputMethods in org.tensorflow.op.nn with parameters of type OperandModifier and TypeMethodDescriptionAvgPool.create(Scope scope, Operand<T> value, List<Long> ksize, List<Long> strides, String padding, AvgPool.Options... options) Factory method to create a class wrapping a new AvgPool operation.AvgPool3d.create(Scope scope, Operand<T> input, List<Long> ksize, List<Long> strides, String padding, AvgPool3d.Options... options) Factory method to create a class wrapping a new AvgPool3D operation.static <T extends TNumber>
AvgPool3dGrad<T> AvgPool3dGrad.create(Scope scope, Operand<TInt32> origInputShape, Operand<T> grad, List<Long> ksize, List<Long> strides, String padding, AvgPool3dGrad.Options... options) Factory method to create a class wrapping a new AvgPool3DGrad operation.static <T extends TNumber>
AvgPoolGrad<T> AvgPoolGrad.create(Scope scope, Operand<TInt32> origInputShape, Operand<T> grad, List<Long> ksize, List<Long> strides, String padding, AvgPoolGrad.Options... options) Factory method to create a class wrapping a new AvgPoolGrad operation.static <T extends TType>
BatchNormWithGlobalNormalization<T> BatchNormWithGlobalNormalization.create(Scope scope, Operand<T> t, Operand<T> m, Operand<T> v, Operand<T> beta, Operand<T> gamma, Float varianceEpsilon, Boolean scaleAfterNormalization) Factory method to create a class wrapping a new BatchNormWithGlobalNormalization operation.static <T extends TType>
BatchNormWithGlobalNormalizationGrad<T> BatchNormWithGlobalNormalizationGrad.create(Scope scope, Operand<T> t, Operand<T> m, Operand<T> v, Operand<T> gamma, Operand<T> backprop, Float varianceEpsilon, Boolean scaleAfterNormalization) Factory method to create a class wrapping a new BatchNormWithGlobalNormalizationGrad operation.BiasAdd.create(Scope scope, Operand<T> value, Operand<T> bias, BiasAdd.Options... options) Factory method to create a class wrapping a new BiasAdd operation.static <T extends TType>
BiasAddGrad<T> BiasAddGrad.create(Scope scope, Operand<T> outBackprop, BiasAddGrad.Options... options) Factory method to create a class wrapping a new BiasAddGrad operation.BlockLSTM.create(Scope scope, Operand<TInt64> seqLenMax, Operand<T> x, Operand<T> csPrev, Operand<T> hPrev, Operand<T> w, Operand<T> wci, Operand<T> wcf, Operand<T> wco, Operand<T> b, BlockLSTM.Options... options) Factory method to create a class wrapping a new BlockLSTMV2 operation.static <T extends TNumber>
BlockLSTMGrad<T> BlockLSTMGrad.create(Scope scope, Operand<TInt64> seqLenMax, Operand<T> x, Operand<T> csPrev, Operand<T> hPrev, Operand<T> w, Operand<T> wci, Operand<T> wcf, Operand<T> wco, Operand<T> b, Operand<T> i, Operand<T> cs, Operand<T> f, Operand<T> o, Operand<T> ci, Operand<T> co, Operand<T> h, Operand<T> csGrad, Operand<T> hGrad, Boolean usePeephole) Factory method to create a class wrapping a new BlockLSTMGradV2 operation.static ComputeAccidentalHitsComputeAccidentalHits.create(Scope scope, Operand<TInt64> trueClasses, Operand<TInt64> sampledCandidates, Long numTrue, ComputeAccidentalHits.Options... options) Factory method to create a class wrapping a new ComputeAccidentalHits operation.Conv.create(Scope scope, Operand<T> input, Operand<T> filter, List<Long> strides, String padding, Conv.Options... options) Factory method to create a class wrapping a new Conv operation.Conv2d.create(Scope scope, Operand<T> input, Operand<T> filter, List<Long> strides, String padding, Conv2d.Options... options) Factory method to create a class wrapping a new Conv2D operation.static <T extends TNumber>
Conv2dBackpropFilter<T> Conv2dBackpropFilter.create(Scope scope, Operand<T> input, Operand<TInt32> filterSizes, Operand<T> outBackprop, List<Long> strides, String padding, Conv2dBackpropFilter.Options... options) Factory method to create a class wrapping a new Conv2DBackpropFilter operation.static <T extends TNumber>
Conv2dBackpropFilterV2<T> Conv2dBackpropFilterV2.create(Scope scope, Operand<T> input, Operand<T> filter, Operand<T> outBackprop, List<Long> strides, String padding, Conv2dBackpropFilterV2.Options... options) Factory method to create a class wrapping a new Conv2DBackpropFilterV2 operation.static <T extends TNumber>
Conv2dBackpropInput<T> Conv2dBackpropInput.create(Scope scope, Operand<TInt32> inputSizes, Operand<T> filter, Operand<T> outBackprop, List<Long> strides, String padding, Conv2dBackpropInput.Options... options) Factory method to create a class wrapping a new Conv2DBackpropInput operation.static <T extends TNumber>
Conv2dBackpropInputV2<T> Conv2dBackpropInputV2.create(Scope scope, Operand<T> input, Operand<T> filter, Operand<T> outBackprop, List<Long> strides, String padding, Conv2dBackpropInputV2.Options... options) Factory method to create a class wrapping a new Conv2DBackpropInputV2 operation.Conv3d.create(Scope scope, Operand<T> input, Operand<T> filter, List<Long> strides, String padding, Conv3d.Options... options) Factory method to create a class wrapping a new Conv3D operation.static <T extends TNumber>
Conv3dBackpropFilter<T> Conv3dBackpropFilter.create(Scope scope, Operand<T> input, Operand<TInt32> filterSizes, Operand<T> outBackprop, List<Long> strides, String padding, Conv3dBackpropFilter.Options... options) Factory method to create a class wrapping a new Conv3DBackpropFilterV2 operation.static <U extends TNumber>
Conv3dBackpropInput<U> Conv3dBackpropInput.create(Scope scope, Operand<? extends TNumber> inputSizes, Operand<U> filter, Operand<U> outBackprop, List<Long> strides, String padding, Conv3dBackpropInput.Options... options) Factory method to create a class wrapping a new Conv3DBackpropInputV2 operation.static <T extends TNumber>
CtcBeamSearchDecoder<T> CtcBeamSearchDecoder.create(Scope scope, Operand<T> inputs, Operand<TInt32> sequenceLength, Long beamWidth, Long topPaths, CtcBeamSearchDecoder.Options... options) Factory method to create a class wrapping a new CTCBeamSearchDecoder operation.static <T extends TNumber>
CtcGreedyDecoder<T> CtcGreedyDecoder.create(Scope scope, Operand<T> inputs, Operand<TInt32> sequenceLength, CtcGreedyDecoder.Options... options) Factory method to create a class wrapping a new CTCGreedyDecoder operation.CtcLoss.create(Scope scope, Operand<T> inputs, Operand<TInt64> labelsIndices, Operand<TInt32> labelsValues, Operand<TInt32> sequenceLength, CtcLoss.Options... options) Factory method to create a class wrapping a new CTCLoss operation.static CTCLossV2CTCLossV2.create(Scope scope, Operand<TFloat32> inputs, Operand<TInt64> labelsIndices, Operand<TInt32> labelsValues, Operand<TInt32> sequenceLength, CTCLossV2.Options... options) Factory method to create a class wrapping a new CTCLossV2 operation.CudnnRNN.create(Scope scope, Operand<T> input, Operand<T> inputH, Operand<T> inputC, Operand<T> params, Operand<TInt32> sequenceLengths, CudnnRNN.Options... options) Factory method to create a class wrapping a new CudnnRNNV3 operation.static <T extends TNumber>
CudnnRNNBackprop<T> CudnnRNNBackprop.create(Scope scope, Operand<T> input, Operand<T> inputH, Operand<T> inputC, Operand<T> params, Operand<TInt32> sequenceLengths, Operand<T> output, Operand<T> outputH, Operand<T> outputC, Operand<T> outputBackprop, Operand<T> outputHBackprop, Operand<T> outputCBackprop, Operand<T> reserveSpace, Operand<? extends TType> hostReserved, CudnnRNNBackprop.Options... options) Factory method to create a class wrapping a new CudnnRNNBackpropV3 operation.static <T extends TNumber>
CudnnRNNCanonicalToParams<T> CudnnRNNCanonicalToParams.create(Scope scope, Operand<TInt32> numLayers, Operand<TInt32> numUnits, Operand<TInt32> inputSize, Iterable<Operand<T>> weights, Iterable<Operand<T>> biases, CudnnRNNCanonicalToParams.Options... options) Factory method to create a class wrapping a new CudnnRNNCanonicalToParamsV2 operation.static <T extends TNumber, U extends TNumber>
CudnnRnnParamsSize<T> CudnnRnnParamsSize.create(Scope scope, Operand<TInt32> numLayers, Operand<TInt32> numUnits, Operand<TInt32> inputSize, Class<U> T, Class<T> S, CudnnRnnParamsSize.Options... options) Factory method to create a class wrapping a new CudnnRNNParamsSize operation.static <T extends TNumber>
CudnnRNNParamsToCanonical<T> CudnnRNNParamsToCanonical.create(Scope scope, Operand<TInt32> numLayers, Operand<TInt32> numUnits, Operand<TInt32> inputSize, Operand<T> params, Long numParamsWeights, Long numParamsBiases, CudnnRNNParamsToCanonical.Options... options) Factory method to create a class wrapping a new CudnnRNNParamsToCanonicalV2 operation.static <T extends TNumber>
DataFormatDimMap<T> DataFormatDimMap.create(Scope scope, Operand<T> x, DataFormatDimMap.Options... options) Factory method to create a class wrapping a new DataFormatDimMap operation.static <T extends TNumber>
DataFormatVecPermute<T> DataFormatVecPermute.create(Scope scope, Operand<T> x, DataFormatVecPermute.Options... options) Factory method to create a class wrapping a new DataFormatVecPermute operation.static <T extends TType>
DepthToSpace<T> DepthToSpace.create(Scope scope, Operand<T> input, Long blockSize, DepthToSpace.Options... options) Factory method to create a class wrapping a new DepthToSpace operation.static <T extends TNumber>
DepthwiseConv2dNative<T> DepthwiseConv2dNative.create(Scope scope, Operand<T> input, Operand<T> filter, List<Long> strides, String padding, DepthwiseConv2dNative.Options... options) Factory method to create a class wrapping a new DepthwiseConv2dNative operation.static <T extends TNumber>
DepthwiseConv2dNativeBackpropFilter<T> DepthwiseConv2dNativeBackpropFilter.create(Scope scope, Operand<T> input, Operand<TInt32> filterSizes, Operand<T> outBackprop, List<Long> strides, String padding, DepthwiseConv2dNativeBackpropFilter.Options... options) Factory method to create a class wrapping a new DepthwiseConv2dNativeBackpropFilter operation.static <T extends TNumber>
DepthwiseConv2dNativeBackpropInput<T> DepthwiseConv2dNativeBackpropInput.create(Scope scope, Operand<TInt32> inputSizes, Operand<T> filter, Operand<T> outBackprop, List<Long> strides, String padding, DepthwiseConv2dNativeBackpropInput.Options... options) Factory method to create a class wrapping a new DepthwiseConv2dNativeBackpropInput operation.static <T extends TNumber>
Dilation2d<T> Dilation2d.create(Scope scope, Operand<T> input, Operand<T> filter, List<Long> strides, List<Long> rates, String padding) Factory method to create a class wrapping a new Dilation2D operation.static <T extends TNumber>
Dilation2dBackpropFilter<T> Dilation2dBackpropFilter.create(Scope scope, Operand<T> input, Operand<T> filter, Operand<T> outBackprop, List<Long> strides, List<Long> rates, String padding) Factory method to create a class wrapping a new Dilation2DBackpropFilter operation.static <T extends TNumber>
Dilation2dBackpropInput<T> Dilation2dBackpropInput.create(Scope scope, Operand<T> input, Operand<T> filter, Operand<T> outBackprop, List<Long> strides, List<Long> rates, String padding) Factory method to create a class wrapping a new Dilation2DBackpropInput operation.Factory method to create a class wrapping a new Elu operation.Factory method to create a class wrapping a new EluGrad operation.static FixedUnigramCandidateSamplerFixedUnigramCandidateSampler.create(Scope scope, Operand<TInt64> trueClasses, Long numTrue, Long numSampled, Boolean unique, Long rangeMax, FixedUnigramCandidateSampler.Options... options) Factory method to create a class wrapping a new FixedUnigramCandidateSampler operation.static <T extends TNumber>
FractionalAvgPool<T> FractionalAvgPool.create(Scope scope, Operand<T> value, List<Float> poolingRatio, FractionalAvgPool.Options... options) Factory method to create a class wrapping a new FractionalAvgPool operation.static <T extends TNumber>
FractionalAvgPoolGrad<T> FractionalAvgPoolGrad.create(Scope scope, Operand<TInt64> origInputTensorShape, Operand<T> outBackprop, Operand<TInt64> rowPoolingSequence, Operand<TInt64> colPoolingSequence, FractionalAvgPoolGrad.Options... options) Factory method to create a class wrapping a new FractionalAvgPoolGrad operation.static <T extends TNumber>
FractionalMaxPool<T> FractionalMaxPool.create(Scope scope, Operand<T> value, List<Float> poolingRatio, FractionalMaxPool.Options... options) Factory method to create a class wrapping a new FractionalMaxPool operation.static <T extends TNumber>
FractionalMaxPoolGrad<T> FractionalMaxPoolGrad.create(Scope scope, Operand<T> origInput, Operand<T> origOutput, Operand<T> outBackprop, Operand<TInt64> rowPoolingSequence, Operand<TInt64> colPoolingSequence, FractionalMaxPoolGrad.Options... options) Factory method to create a class wrapping a new FractionalMaxPoolGrad operation.static <T extends TNumber, U extends TNumber>
FusedBatchNorm<T, U> FusedBatchNorm.create(Scope scope, Operand<T> x, Operand<U> scale, Operand<U> offset, Operand<U> mean, Operand<U> variance, FusedBatchNorm.Options... options) Factory method to create a class wrapping a new FusedBatchNormV3 operation.static <T extends TNumber, U extends TNumber>
FusedBatchNormGrad<T, U> FusedBatchNormGrad.create(Scope scope, Operand<T> yBackprop, Operand<T> x, Operand<TFloat32> scale, Operand<U> reserveSpace1, Operand<U> reserveSpace2, Operand<U> reserveSpace3, FusedBatchNormGrad.Options... options) Factory method to create a class wrapping a new FusedBatchNormGradV3 operation.static <T extends TNumber>
FusedPadConv2d<T> FusedPadConv2d.create(Scope scope, Operand<T> input, Operand<TInt32> paddings, Operand<T> filter, String mode, List<Long> strides, String padding) Factory method to create a class wrapping a new FusedPadConv2D operation.static <T extends TNumber>
FusedResizeAndPadConv2d<T> FusedResizeAndPadConv2d.create(Scope scope, Operand<T> input, Operand<TInt32> sizeOutput, Operand<TInt32> paddings, Operand<T> filter, String mode, List<Long> strides, String padding, FusedResizeAndPadConv2d.Options... options) Factory method to create a class wrapping a new FusedResizeAndPadConv2D operation.static <T extends TNumber>
GRUBlockCell<T> GRUBlockCell.create(Scope scope, Operand<T> x, Operand<T> hPrev, Operand<T> wRu, Operand<T> wC, Operand<T> bRu, Operand<T> bC) Factory method to create a class wrapping a new GRUBlockCell operation.static <T extends TNumber>
GRUBlockCellGrad<T> GRUBlockCellGrad.create(Scope scope, Operand<T> x, Operand<T> hPrev, Operand<T> wRu, Operand<T> wC, Operand<T> bRu, Operand<T> bC, Operand<T> r, Operand<T> u, Operand<T> c, Operand<T> dH) Factory method to create a class wrapping a new GRUBlockCellGrad operation.Factory method to create a class wrapping a new InTopKV2 operation.Factory method to create a class wrapping a new InvGrad operation.static IsotonicRegression<TFloat32> Factory method to create a class wrapping a new IsotonicRegression operation, with the default output types.static <U extends TNumber>
IsotonicRegression<U> Factory method to create a class wrapping a new IsotonicRegression operation.Factory method to create a class wrapping a new L2Loss operation.LeakyRelu.create(Scope scope, Operand<T> features, LeakyRelu.Options... options) Factory method to create a class wrapping a new LeakyRelu operation.LearnedUnigramCandidateSampler.create(Scope scope, Operand<TInt64> trueClasses, Long numTrue, Long numSampled, Boolean unique, Long rangeMax, LearnedUnigramCandidateSampler.Options... options) Factory method to create a class wrapping a new LearnedUnigramCandidateSampler operation.static <T extends TNumber>
LocalResponseNormalization<T> LocalResponseNormalization.create(Scope scope, Operand<T> input, LocalResponseNormalization.Options... options) Factory method to create a class wrapping a new LRN operation.static <T extends TNumber>
LocalResponseNormalizationGrad<T> LocalResponseNormalizationGrad.create(Scope scope, Operand<T> inputGrads, Operand<T> inputImage, Operand<T> outputImage, LocalResponseNormalizationGrad.Options... options) Factory method to create a class wrapping a new LRNGrad operation.static <T extends TNumber>
LogSoftmax<T> Factory method to create a class wrapping a new LogSoftmax operation.static <T extends TNumber>
LSTMBlockCell<T> LSTMBlockCell.create(Scope scope, Operand<T> x, Operand<T> csPrev, Operand<T> hPrev, Operand<T> w, Operand<T> wci, Operand<T> wcf, Operand<T> wco, Operand<T> b, LSTMBlockCell.Options... options) Factory method to create a class wrapping a new LSTMBlockCell operation.static <T extends TNumber>
LSTMBlockCellGrad<T> LSTMBlockCellGrad.create(Scope scope, Operand<T> x, Operand<T> csPrev, Operand<T> hPrev, Operand<T> w, Operand<T> wci, Operand<T> wcf, Operand<T> wco, Operand<T> b, Operand<T> i, Operand<T> cs, Operand<T> f, Operand<T> o, Operand<T> ci, Operand<T> co, Operand<T> csGrad, Operand<T> hGrad, Boolean usePeephole) Factory method to create a class wrapping a new LSTMBlockCellGrad operation.MaxPool.create(Scope scope, Operand<T> input, Operand<TInt32> ksize, Operand<TInt32> strides, String padding, MaxPool.Options... options) Factory method to create a class wrapping a new MaxPoolV2 operation.MaxPool3d.create(Scope scope, Operand<T> input, List<Long> ksize, List<Long> strides, String padding, MaxPool3d.Options... options) Factory method to create a class wrapping a new MaxPool3D operation.static <U extends TNumber, T extends TNumber>
MaxPool3dGrad<U> MaxPool3dGrad.create(Scope scope, Operand<T> origInput, Operand<T> origOutput, Operand<U> grad, List<Long> ksize, List<Long> strides, String padding, MaxPool3dGrad.Options... options) Factory method to create a class wrapping a new MaxPool3DGrad operation.static <T extends TNumber>
MaxPool3dGradGrad<T> MaxPool3dGradGrad.create(Scope scope, Operand<T> origInput, Operand<T> origOutput, Operand<T> grad, List<Long> ksize, List<Long> strides, String padding, MaxPool3dGradGrad.Options... options) Factory method to create a class wrapping a new MaxPool3DGradGrad operation.static <T extends TNumber>
MaxPoolGrad<T> MaxPoolGrad.create(Scope scope, Operand<T> origInput, Operand<T> origOutput, Operand<T> grad, Operand<TInt32> ksize, Operand<TInt32> strides, String padding, MaxPoolGrad.Options... options) Factory method to create a class wrapping a new MaxPoolGradV2 operation.static <T extends TNumber>
MaxPoolGradGrad<T> MaxPoolGradGrad.create(Scope scope, Operand<T> origInput, Operand<T> origOutput, Operand<T> grad, Operand<TInt32> ksize, Operand<TInt32> strides, String padding, MaxPoolGradGrad.Options... options) Factory method to create a class wrapping a new MaxPoolGradGradV2 operation.static <T extends TNumber>
MaxPoolGradGradWithArgmax<T> MaxPoolGradGradWithArgmax.create(Scope scope, Operand<T> input, Operand<T> grad, Operand<? extends TNumber> argmax, List<Long> ksize, List<Long> strides, String padding, MaxPoolGradGradWithArgmax.Options... options) Factory method to create a class wrapping a new MaxPoolGradGradWithArgmax operation.static <T extends TNumber>
MaxPoolGradWithArgmax<T> MaxPoolGradWithArgmax.create(Scope scope, Operand<T> input, Operand<T> grad, Operand<? extends TNumber> argmax, List<Long> ksize, List<Long> strides, String padding, MaxPoolGradWithArgmax.Options... options) Factory method to create a class wrapping a new MaxPoolGradWithArgmax operation.static <T extends TNumber, U extends TNumber>
MaxPoolWithArgmax<T, U> MaxPoolWithArgmax.create(Scope scope, Operand<T> input, List<Long> ksize, List<Long> strides, Class<U> Targmax, String padding, MaxPoolWithArgmax.Options... options) Factory method to create a class wrapping a new MaxPoolWithArgmax operation.static <T extends TNumber>
MaxPoolWithArgmax<T, TInt64> MaxPoolWithArgmax.create(Scope scope, Operand<T> input, List<Long> ksize, List<Long> strides, String padding, MaxPoolWithArgmax.Options[] options) Factory method to create a class wrapping a new MaxPoolWithArgmax operation, with the default output types.static <T extends TNumber>
NthElement<T> Factory method to create a class wrapping a new NthElement operation.static <T extends TNumber>
QuantizedAvgPool<T> QuantizedAvgPool.create(Scope scope, Operand<T> input, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, List<Long> ksize, List<Long> strides, String padding) Factory method to create a class wrapping a new QuantizedAvgPool operation.static <U extends TNumber, T extends TNumber>
QuantizedBatchNormWithGlobalNormalization<U> QuantizedBatchNormWithGlobalNormalization.create(Scope scope, Operand<T> t, Operand<TFloat32> tMin, Operand<TFloat32> tMax, Operand<T> m, Operand<TFloat32> mMin, Operand<TFloat32> mMax, Operand<T> v, Operand<TFloat32> vMin, Operand<TFloat32> vMax, Operand<T> beta, Operand<TFloat32> betaMin, Operand<TFloat32> betaMax, Operand<T> gamma, Operand<TFloat32> gammaMin, Operand<TFloat32> gammaMax, Class<U> outType, Float varianceEpsilon, Boolean scaleAfterNormalization) Factory method to create a class wrapping a new QuantizedBatchNormWithGlobalNormalization operation.static <V extends TNumber>
QuantizedBiasAdd<V> QuantizedBiasAdd.create(Scope scope, Operand<? extends TNumber> input, Operand<? extends TNumber> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minBias, Operand<TFloat32> maxBias, Class<V> outType) Factory method to create a class wrapping a new QuantizedBiasAdd operation.static <V extends TNumber>
QuantizedConv2d<V> QuantizedConv2d.create(Scope scope, Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Class<V> outType, List<Long> strides, String padding, QuantizedConv2d.Options... options) Factory method to create a class wrapping a new QuantizedConv2D operation.static <V extends TNumber>
QuantizedConv2DAndRelu<V> QuantizedConv2DAndRelu.create(Scope scope, Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Class<V> outType, List<Long> strides, String padding, QuantizedConv2DAndRelu.Options... options) Factory method to create a class wrapping a new QuantizedConv2DAndRelu operation.static <V extends TNumber>
QuantizedConv2DAndReluAndRequantize<V> QuantizedConv2DAndReluAndRequantize.create(Scope scope, Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Class<V> outType, List<Long> strides, String padding, QuantizedConv2DAndReluAndRequantize.Options... options) Factory method to create a class wrapping a new QuantizedConv2DAndReluAndRequantize operation.static <V extends TNumber>
QuantizedConv2DAndRequantize<V> QuantizedConv2DAndRequantize.create(Scope scope, Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Class<V> outType, List<Long> strides, String padding, QuantizedConv2DAndRequantize.Options... options) Factory method to create a class wrapping a new QuantizedConv2DAndRequantize operation.static <V extends TNumber>
QuantizedConv2DPerChannel<V> QuantizedConv2DPerChannel.create(Scope scope, Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Class<V> outType, List<Long> strides, String padding, QuantizedConv2DPerChannel.Options... options) Factory method to create a class wrapping a new QuantizedConv2DPerChannel operation.static <V extends TNumber>
QuantizedConv2DWithBias<V> QuantizedConv2DWithBias.create(Scope scope, Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Class<V> outType, List<Long> strides, String padding, QuantizedConv2DWithBias.Options... options) Factory method to create a class wrapping a new QuantizedConv2DWithBias operation.static <V extends TNumber>
QuantizedConv2DWithBiasAndRelu<V> QuantizedConv2DWithBiasAndRelu.create(Scope scope, Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Class<V> outType, List<Long> strides, String padding, QuantizedConv2DWithBiasAndRelu.Options... options) Factory method to create a class wrapping a new QuantizedConv2DWithBiasAndRelu operation.static <W extends TNumber>
QuantizedConv2DWithBiasAndReluAndRequantize<W> QuantizedConv2DWithBiasAndReluAndRequantize.create(Scope scope, Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<? extends TNumber> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Class<W> outType, List<Long> strides, String padding, QuantizedConv2DWithBiasAndReluAndRequantize.Options... options) Factory method to create a class wrapping a new QuantizedConv2DWithBiasAndReluAndRequantize operation.static <W extends TNumber>
QuantizedConv2DWithBiasAndRequantize<W> QuantizedConv2DWithBiasAndRequantize.create(Scope scope, Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<? extends TNumber> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Class<W> outType, List<Long> strides, String padding, QuantizedConv2DWithBiasAndRequantize.Options... options) Factory method to create a class wrapping a new QuantizedConv2DWithBiasAndRequantize operation.static <X extends TNumber>
QuantizedConv2DWithBiasSignedSumAndReluAndRequantize<X> QuantizedConv2DWithBiasSignedSumAndReluAndRequantize.create(Scope scope, Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<? extends TNumber> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Operand<? extends TNumber> summand, Operand<TFloat32> minSummand, Operand<TFloat32> maxSummand, Class<X> outType, List<Long> strides, String padding, QuantizedConv2DWithBiasSignedSumAndReluAndRequantize.Options... options) Factory method to create a class wrapping a new QuantizedConv2DWithBiasSignedSumAndReluAndRequantize operation.static <V extends TNumber>
QuantizedConv2DWithBiasSumAndRelu<V> QuantizedConv2DWithBiasSumAndRelu.create(Scope scope, Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Operand<TFloat32> summand, Class<V> outType, List<Long> strides, String padding, QuantizedConv2DWithBiasSumAndRelu.Options... options) Factory method to create a class wrapping a new QuantizedConv2DWithBiasSumAndRelu operation.static <X extends TNumber>
QuantizedConv2DWithBiasSumAndReluAndRequantize<X> QuantizedConv2DWithBiasSumAndReluAndRequantize.create(Scope scope, Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<? extends TNumber> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Operand<? extends TNumber> summand, Operand<TFloat32> minSummand, Operand<TFloat32> maxSummand, Class<X> outType, List<Long> strides, String padding, QuantizedConv2DWithBiasSumAndReluAndRequantize.Options... options) Factory method to create a class wrapping a new QuantizedConv2DWithBiasSumAndReluAndRequantize operation.static <V extends TNumber>
QuantizedDepthwiseConv2D<V> QuantizedDepthwiseConv2D.create(Scope scope, Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Class<V> outType, List<Long> strides, String padding, QuantizedDepthwiseConv2D.Options... options) Factory method to create a class wrapping a new QuantizedDepthwiseConv2D operation.static <V extends TNumber>
QuantizedDepthwiseConv2DWithBias<V> QuantizedDepthwiseConv2DWithBias.create(Scope scope, Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Class<V> outType, List<Long> strides, String padding, QuantizedDepthwiseConv2DWithBias.Options... options) Factory method to create a class wrapping a new QuantizedDepthwiseConv2DWithBias operation.static <V extends TNumber>
QuantizedDepthwiseConv2DWithBiasAndRelu<V> QuantizedDepthwiseConv2DWithBiasAndRelu.create(Scope scope, Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<TFloat32> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Class<V> outType, List<Long> strides, String padding, QuantizedDepthwiseConv2DWithBiasAndRelu.Options... options) Factory method to create a class wrapping a new QuantizedDepthwiseConv2DWithBiasAndRelu operation.static <W extends TNumber>
QuantizedDepthwiseConv2DWithBiasAndReluAndRequantize<W> QuantizedDepthwiseConv2DWithBiasAndReluAndRequantize.create(Scope scope, Operand<? extends TNumber> input, Operand<? extends TNumber> filter, Operand<? extends TNumber> bias, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, Operand<TFloat32> minFilter, Operand<TFloat32> maxFilter, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Class<W> outType, List<Long> strides, String padding, QuantizedDepthwiseConv2DWithBiasAndReluAndRequantize.Options... options) Factory method to create a class wrapping a new QuantizedDepthwiseConv2DWithBiasAndReluAndRequantize operation.static <T extends TNumber>
QuantizedInstanceNorm<T> QuantizedInstanceNorm.create(Scope scope, Operand<T> x, Operand<TFloat32> xMin, Operand<TFloat32> xMax, QuantizedInstanceNorm.Options... options) Factory method to create a class wrapping a new QuantizedInstanceNorm operation.static <T extends TNumber>
QuantizedMaxPool<T> QuantizedMaxPool.create(Scope scope, Operand<T> input, Operand<TFloat32> minInput, Operand<TFloat32> maxInput, List<Long> ksize, List<Long> strides, String padding) Factory method to create a class wrapping a new QuantizedMaxPool operation.static <U extends TNumber>
QuantizedRelu<U> QuantizedRelu.create(Scope scope, Operand<? extends TNumber> features, Operand<TFloat32> minFeatures, Operand<TFloat32> maxFeatures, Class<U> outType) Factory method to create a class wrapping a new QuantizedRelu operation.static <U extends TNumber>
QuantizedRelu6<U> QuantizedRelu6.create(Scope scope, Operand<? extends TNumber> features, Operand<TFloat32> minFeatures, Operand<TFloat32> maxFeatures, Class<U> outType) Factory method to create a class wrapping a new QuantizedRelu6 operation.static <U extends TNumber>
QuantizedReluX<U> QuantizedReluX.create(Scope scope, Operand<? extends TNumber> features, Operand<TFloat32> maxValue, Operand<TFloat32> minFeatures, Operand<TFloat32> maxFeatures, Class<U> outType) Factory method to create a class wrapping a new QuantizedReluX operation.Factory method to create a class wrapping a new Relu operation.Factory method to create a class wrapping a new Relu6 operation.Factory method to create a class wrapping a new Relu6Grad operation.Factory method to create a class wrapping a new ReluGrad operation.Factory method to create a class wrapping a new Selu operation.Factory method to create a class wrapping a new SeluGrad operation.Factory method to create a class wrapping a new Softmax operation.static <T extends TNumber>
SoftmaxCrossEntropyWithLogits<T> Factory method to create a class wrapping a new SoftmaxCrossEntropyWithLogits operation.Factory method to create a class wrapping a new Softsign operation.static <T extends TNumber>
SoftsignGrad<T> Factory method to create a class wrapping a new SoftsignGrad operation.static <T extends TType>
SpaceToBatch<T> SpaceToBatch.create(Scope scope, Operand<T> input, Operand<? extends TNumber> paddings, Long blockSize) Factory method to create a class wrapping a new SpaceToBatch operation.static <T extends TType>
SpaceToDepth<T> SpaceToDepth.create(Scope scope, Operand<T> input, Long blockSize, SpaceToDepth.Options... options) Factory method to create a class wrapping a new SpaceToDepth operation.static <T extends TNumber>
SparseSoftmaxCrossEntropyWithLogits<T> SparseSoftmaxCrossEntropyWithLogits.create(Scope scope, Operand<T> features, Operand<? extends TNumber> labels) Factory method to create a class wrapping a new SparseSoftmaxCrossEntropyWithLogits operation.TopK.create(Scope scope, Operand<T> input, Operand<? extends TNumber> k, Class<V> indexType, TopK.Options... options) Factory method to create a class wrapping a new TopKV2 operation.Factory method to create a class wrapping a new TopKV2 operation, with the default output types.static <U extends TNumber, T extends TNumber>
UniformQuantizedConvolution<U> UniformQuantizedConvolution.create(Scope scope, Operand<T> lhs, Operand<T> rhs, Operand<TFloat32> lhsScales, Operand<TInt32> lhsZeroPoints, Operand<TFloat32> rhsScales, Operand<TInt32> rhsZeroPoints, Operand<TFloat32> outputScales, Operand<TInt32> outputZeroPoints, Class<U> Tout, String padding, Long lhsQuantizationMinVal, Long lhsQuantizationMaxVal, Long rhsQuantizationMinVal, Long rhsQuantizationMaxVal, Long outputQuantizationMinVal, Long outputQuantizationMaxVal, UniformQuantizedConvolution.Options... options) Factory method to create a class wrapping a new UniformQuantizedConvolution operation.static <V extends TNumber>
UniformQuantizedConvolutionHybrid<V> UniformQuantizedConvolutionHybrid.create(Scope scope, Operand<? extends TNumber> lhs, Operand<? extends TNumber> rhs, Operand<TFloat32> rhsScales, Operand<TInt32> rhsZeroPoints, Class<V> Tout, String padding, Long rhsQuantizationMinVal, Long rhsQuantizationMaxVal, UniformQuantizedConvolutionHybrid.Options... options) Factory method to create a class wrapping a new UniformQuantizedConvolutionHybrid operation.Method parameters in org.tensorflow.op.nn with type arguments of type OperandModifier and TypeMethodDescriptionstatic <T extends TNumber>
CudnnRNNCanonicalToParams<T> CudnnRNNCanonicalToParams.create(Scope scope, Operand<TInt32> numLayers, Operand<TInt32> numUnits, Operand<TInt32> inputSize, Iterable<Operand<T>> weights, Iterable<Operand<T>> biases, CudnnRNNCanonicalToParams.Options... options) Factory method to create a class wrapping a new CudnnRNNCanonicalToParamsV2 operation. -
Uses of Operand in org.tensorflow.op.quantization
Classes in org.tensorflow.op.quantization that implement OperandModifier and TypeClassDescriptionfinal classDequantize<U extends TNumber>Dequantize the 'input' tensor into a float or bfloat16 Tensor.final classFake-quantize the 'inputs' tensor, type float to 'outputs' tensor of same shape and type.final classCompute gradients for a FakeQuantWithMinMaxArgs operation.final classFake-quantize the 'inputs' tensor of type float via global float scalars Fake-quantize theinputstensor of type float via global float scalarsminandmaxtooutputstensor of same shape asinputs.final classFake-quantize the 'inputs' tensor of type float via per-channel floats Fake-quantize theinputstensor of type float per-channel and one of the shapes:[d],[b, d][b, h, w, d]via per-channel floatsminandmaxof shape[d]tooutputstensor of same shape asinputs.final classQuantizeAndDequantize<T extends TNumber>Quantizes then dequantizes a tensor.final classQuantizeAndDequantizeV3<T extends TNumber>Quantizes then dequantizes a tensor.final classQuantizeAndDequantizeV4<T extends TNumber>Quantizes then dequantizes a tensor.final classQuantizedMatMulWithBiasAndDequantize<W extends TNumber>The QuantizedMatMulWithBiasAndDequantize operationfinal classUniformDequantize<U extends TNumber>Perform dequantization on the quantized Tensorinput.final classUniformQuantize<U extends TNumber>Perform quantization on Tensorinput.final classUniformQuantizedDot<U extends TNumber>Perform quantized dot of quantized Tensorlhsand quantized Tensorrhsto make quantizedoutput.final classUniformQuantizedDotHybrid<V extends TNumber>Perform hybrid quantized dot of float Tensorlhsand quantized Tensorrhs.final classUniformRequantize<U extends TNumber>Given quantized tensorinput, requantize it with new quantization parameters.Fields in org.tensorflow.op.quantization declared as OperandModifier and TypeFieldDescriptionQuantizedMatMulWithBiasAndDequantize.Inputs.aThe a inputQuantizedMatMulWithBiasAndRequantize.Inputs.aThe a inputQuantizedMatMulWithBiasAndDequantize.Inputs.bThe b inputQuantizedMatMulWithBiasAndRequantize.Inputs.bThe b inputQuantizedMatMulWithBiasAndDequantize.Inputs.biasThe bias inputQuantizedMatMulWithBiasAndRequantize.Inputs.biasThe bias inputQuantizedConcat.Inputs.concatDim0-D.FakeQuantWithMinMaxArgsGradient.Inputs.gradientsBackpropagated gradients above the FakeQuantWithMinMaxArgs operation.FakeQuantWithMinMaxVarsGradient.Inputs.gradientsBackpropagated gradients above the FakeQuantWithMinMaxVars operation.FakeQuantWithMinMaxVarsPerChannelGradient.Inputs.gradientsBackpropagated gradients above the FakeQuantWithMinMaxVars operation, shape one of:[d],[b, d],[b, h, w, d].QuantizeAndDequantizeV4Grad.Inputs.gradientsThe gradients inputDequantize.Inputs.inputThe input inputQuantize.Inputs.inputThe input inputQuantizeAndDequantize.Inputs.inputThe input inputQuantizeAndDequantizeV3.Inputs.inputThe input inputQuantizeAndDequantizeV4.Inputs.inputTensor to quantize and then dequantize.QuantizeAndDequantizeV4Grad.Inputs.inputThe input inputQuantizeDownAndShrinkRange.Inputs.inputThe input inputRequantizationRange.Inputs.inputThe input inputRequantize.Inputs.inputThe input inputUniformDequantize.Inputs.inputMust be a Tensor of Tin.UniformQuantize.Inputs.inputMust be a Tensor of Tin.UniformRequantize.Inputs.inputMust be a Tensor of Tin.QuantizeAndDequantize.Inputs.inputMaxThe inputMax inputQuantizeAndDequantizeV3.Inputs.inputMaxThe inputMax inputQuantizeAndDequantizeV4.Inputs.inputMaxIfrange_given == True, this specifies the maximum input value that needs to be represented, otherwise it is determined from the max value of theinputtensor.QuantizeAndDequantizeV4Grad.Inputs.inputMaxThe inputMax inputQuantizeDownAndShrinkRange.Inputs.inputMaxThe float value that the maximum quantized input value represents.RequantizationRange.Inputs.inputMaxThe float value that the maximum quantized input value represents.Requantize.Inputs.inputMaxThe float value that the maximum quantized input value represents.QuantizeAndDequantize.Inputs.inputMinThe inputMin inputQuantizeAndDequantizeV3.Inputs.inputMinThe inputMin inputQuantizeAndDequantizeV4.Inputs.inputMinIfrange_given == True, this specifies the minimum input value that needs to be represented, otherwise it is determined from the min value of theinputtensor.QuantizeAndDequantizeV4Grad.Inputs.inputMinThe inputMin inputQuantizeDownAndShrinkRange.Inputs.inputMinThe float value that the minimum quantized input value represents.RequantizationRange.Inputs.inputMinThe float value that the minimum quantized input value represents.Requantize.Inputs.inputMinThe float value that the minimum quantized input value represents.FakeQuantWithMinMaxArgs.Inputs.inputsThe inputs inputFakeQuantWithMinMaxArgsGradient.Inputs.inputsValues passed as inputs to the FakeQuantWithMinMaxArgs operation.FakeQuantWithMinMaxVars.Inputs.inputsThe inputs inputFakeQuantWithMinMaxVarsGradient.Inputs.inputsValues passed as inputs to the FakeQuantWithMinMaxVars operation. min, max: Quantization interval, scalar floats.FakeQuantWithMinMaxVarsPerChannel.Inputs.inputsThe inputs inputFakeQuantWithMinMaxVarsPerChannelGradient.Inputs.inputsValues passed as inputs to the FakeQuantWithMinMaxVars operation, shape same asgradients. min, max: Quantization interval, floats of shape[d].UniformRequantize.Inputs.inputScalesThe float value(s) used as scale(s) when quantizing original data thatinputrepresents.UniformRequantize.Inputs.inputZeroPointsThe int32 value(s) used as zero_point(s) when quantizing original data thatinputrepresents.UniformQuantizedDot.Inputs.lhsMust be a 2D Tensor of Tin.UniformQuantizedDotHybrid.Inputs.lhsMust be a 2D Tensor of Tlhs.UniformQuantizedDot.Inputs.lhsScalesThe float value(s) used as scale when quantizing original data that lhs represents.UniformQuantizedDot.Inputs.lhsZeroPointsThe int32 value(s) used as zero_point when quantizing original data that lhs represents.FakeQuantWithMinMaxVars.Inputs.maxThe max inputFakeQuantWithMinMaxVarsGradient.Inputs.maxThe max inputFakeQuantWithMinMaxVarsPerChannel.Inputs.maxThe max inputFakeQuantWithMinMaxVarsPerChannelGradient.Inputs.maxThe max inputQuantizedMatMulWithBiasAndDequantize.Inputs.maxAThe maxA inputQuantizedMatMulWithBiasAndRequantize.Inputs.maxAThe maxA inputQuantizedMatMulWithBiasAndDequantize.Inputs.maxBThe maxB inputQuantizedMatMulWithBiasAndRequantize.Inputs.maxBThe maxB inputQuantizedMatMulWithBiasAndDequantize.Inputs.maxFreezedOutputThe maxFreezedOutput inputQuantizedMatMulWithBiasAndRequantize.Inputs.maxFreezedOutputThe maxFreezedOutput inputDequantize.Inputs.maxRangeThe maximum scalar value possibly produced for the input.Quantize.Inputs.maxRangeThe maximum value of the quantization range.FakeQuantWithMinMaxVars.Inputs.minThe min inputFakeQuantWithMinMaxVarsGradient.Inputs.minThe min inputFakeQuantWithMinMaxVarsPerChannel.Inputs.minThe min inputFakeQuantWithMinMaxVarsPerChannelGradient.Inputs.minThe min inputQuantizedMatMulWithBiasAndDequantize.Inputs.minAThe minA inputQuantizedMatMulWithBiasAndRequantize.Inputs.minAThe minA inputQuantizedMatMulWithBiasAndDequantize.Inputs.minBThe minB inputQuantizedMatMulWithBiasAndRequantize.Inputs.minBThe minB inputQuantizedMatMulWithBiasAndDequantize.Inputs.minFreezedOutputThe minFreezedOutput inputQuantizedMatMulWithBiasAndRequantize.Inputs.minFreezedOutputThe minFreezedOutput inputDequantize.Inputs.minRangeThe minimum scalar value possibly produced for the input.Quantize.Inputs.minRangeThe minimum value of the quantization range.QuantizeAndDequantize.Inputs.numBitsThe numBits inputQuantizeAndDequantizeV3.Inputs.numBitsThe numBits inputUniformQuantizedDot.Inputs.outputScalesThe float value(s) to use as scales when quantizing original data that output represents.UniformRequantize.Inputs.outputScalesThe float value(s) to use as new scale(s) to quantize original data thatinputrepresents.UniformQuantizedDot.Inputs.outputZeroPointsThe int32 value(s) used as zero_point when quantizing original data that output represents.UniformRequantize.Inputs.outputZeroPointsThe int32 value(s) to use as new zero_point(s) to quantize original data thatinputrepresents.Requantize.Inputs.requestedOutputMaxThe float value that the maximum quantized output value represents.Requantize.Inputs.requestedOutputMinThe float value that the minimum quantized output value represents.UniformQuantizedDot.Inputs.rhsMust be a 2D Tensor of Tin.UniformQuantizedDotHybrid.Inputs.rhsMust be a 2D Tensor of Trhs.UniformQuantizedDot.Inputs.rhsScalesThe float value(s) used as scale when quantizing original data that rhs represents.UniformQuantizedDotHybrid.Inputs.rhsScalesThe float value(s) used as scale when quantizing original data that rhs represents.UniformQuantizedDot.Inputs.rhsZeroPointsThe int32 value(s) used as zero_point when quantizing original data that rhs represents.UniformQuantizedDotHybrid.Inputs.rhsZeroPointsThe int32 value(s) used as zero_point when quantizing original data that rhs represents.UniformDequantize.Inputs.scalesThe float value(s) used as scale(s) when quantizing original data that input represents.UniformQuantize.Inputs.scalesThe float value(s) to use as scale(s) to quantizeinput.UniformDequantize.Inputs.zeroPointsThe int32 value(s) used as zero_point(s) when quantizing original data that input represents.UniformQuantize.Inputs.zeroPointsThe int32 value(s) to use as zero_point(s) to quantizeinput.Fields in org.tensorflow.op.quantization with type parameters of type OperandModifier and TypeFieldDescriptionQuantizedConcat.Inputs.inputMaxesThe maximum scalar values for each of the input tensors.QuantizedConcat.Inputs.inputMinsThe minimum scalar values for each of the input tensors.QuantizedConcat.Inputs.valuesTheNTensors to concatenate.Methods in org.tensorflow.op.quantization with parameters of type OperandModifier and TypeMethodDescriptionstatic <U extends TNumber>
Dequantize<U> Dequantize.create(Scope scope, Operand<? extends TNumber> input, Operand<TFloat32> minRange, Operand<TFloat32> maxRange, Class<U> dtype, Dequantize.Options... options) Factory method to create a class wrapping a new Dequantize operation.static Dequantize<TFloat32> Dequantize.create(Scope scope, Operand<? extends TNumber> input, Operand<TFloat32> minRange, Operand<TFloat32> maxRange, Dequantize.Options[] options) Factory method to create a class wrapping a new Dequantize operation, with the default output types.static FakeQuantWithMinMaxArgsFakeQuantWithMinMaxArgs.create(Scope scope, Operand<TFloat32> inputs, FakeQuantWithMinMaxArgs.Options... options) Factory method to create a class wrapping a new FakeQuantWithMinMaxArgs operation.FakeQuantWithMinMaxArgsGradient.create(Scope scope, Operand<TFloat32> gradients, Operand<TFloat32> inputs, FakeQuantWithMinMaxArgsGradient.Options... options) Factory method to create a class wrapping a new FakeQuantWithMinMaxArgsGradient operation.static FakeQuantWithMinMaxVarsFakeQuantWithMinMaxVars.create(Scope scope, Operand<TFloat32> inputs, Operand<TFloat32> min, Operand<TFloat32> max, FakeQuantWithMinMaxVars.Options... options) Factory method to create a class wrapping a new FakeQuantWithMinMaxVars operation.FakeQuantWithMinMaxVarsGradient.create(Scope scope, Operand<TFloat32> gradients, Operand<TFloat32> inputs, Operand<TFloat32> min, Operand<TFloat32> max, FakeQuantWithMinMaxVarsGradient.Options... options) Factory method to create a class wrapping a new FakeQuantWithMinMaxVarsGradient operation.FakeQuantWithMinMaxVarsPerChannel.create(Scope scope, Operand<TFloat32> inputs, Operand<TFloat32> min, Operand<TFloat32> max, FakeQuantWithMinMaxVarsPerChannel.Options... options) Factory method to create a class wrapping a new FakeQuantWithMinMaxVarsPerChannel operation.FakeQuantWithMinMaxVarsPerChannelGradient.create(Scope scope, Operand<TFloat32> gradients, Operand<TFloat32> inputs, Operand<TFloat32> min, Operand<TFloat32> max, FakeQuantWithMinMaxVarsPerChannelGradient.Options... options) Factory method to create a class wrapping a new FakeQuantWithMinMaxVarsPerChannelGradient operation.Quantize.create(Scope scope, Operand<TFloat32> input, Operand<TFloat32> minRange, Operand<TFloat32> maxRange, Class<T> T, Quantize.Options... options) Factory method to create a class wrapping a new QuantizeV2 operation.static <T extends TNumber>
QuantizeAndDequantize<T> QuantizeAndDequantize.create(Scope scope, Operand<T> input, Operand<T> inputMin, Operand<T> inputMax, Operand<TInt32> numBits, QuantizeAndDequantize.Options... options) Factory method to create a class wrapping a new QuantizeAndDequantizeV3 operation.static <T extends TNumber>
QuantizeAndDequantizeV3<T> QuantizeAndDequantizeV3.create(Scope scope, Operand<T> input, Operand<T> inputMin, Operand<T> inputMax, Operand<TInt32> numBits, QuantizeAndDequantizeV3.Options... options) Factory method to create a class wrapping a new QuantizeAndDequantizeV3 operation.static <T extends TNumber>
QuantizeAndDequantizeV4<T> QuantizeAndDequantizeV4.create(Scope scope, Operand<T> input, Operand<T> inputMin, Operand<T> inputMax, QuantizeAndDequantizeV4.Options... options) Factory method to create a class wrapping a new QuantizeAndDequantizeV4 operation.static <T extends TNumber>
QuantizeAndDequantizeV4Grad<T> QuantizeAndDequantizeV4Grad.create(Scope scope, Operand<T> gradients, Operand<T> input, Operand<T> inputMin, Operand<T> inputMax, QuantizeAndDequantizeV4Grad.Options... options) Factory method to create a class wrapping a new QuantizeAndDequantizeV4Grad operation.static <T extends TType>
QuantizedConcat<T> QuantizedConcat.create(Scope scope, Operand<TInt32> concatDim, Iterable<Operand<T>> values, Iterable<Operand<TFloat32>> inputMins, Iterable<Operand<TFloat32>> inputMaxes) Factory method to create a class wrapping a new QuantizedConcat operation.static <W extends TNumber>
QuantizedMatMulWithBiasAndDequantize<W> QuantizedMatMulWithBiasAndDequantize.create(Scope scope, Operand<? extends TNumber> a, Operand<? extends TNumber> b, Operand<? extends TNumber> bias, Operand<TFloat32> minA, Operand<TFloat32> maxA, Operand<TFloat32> minB, Operand<TFloat32> maxB, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Class<W> Toutput, QuantizedMatMulWithBiasAndDequantize.Options... options) Factory method to create a class wrapping a new QuantizedMatMulWithBiasAndDequantize operation.static <W extends TNumber>
QuantizedMatMulWithBiasAndRequantize<W> QuantizedMatMulWithBiasAndRequantize.create(Scope scope, Operand<? extends TNumber> a, Operand<? extends TNumber> b, Operand<? extends TNumber> bias, Operand<TFloat32> minA, Operand<TFloat32> maxA, Operand<TFloat32> minB, Operand<TFloat32> maxB, Operand<TFloat32> minFreezedOutput, Operand<TFloat32> maxFreezedOutput, Class<W> Toutput, QuantizedMatMulWithBiasAndRequantize.Options... options) Factory method to create a class wrapping a new QuantizedMatMulWithBiasAndRequantize operation.static <U extends TNumber>
QuantizeDownAndShrinkRange<U> QuantizeDownAndShrinkRange.create(Scope scope, Operand<? extends TNumber> input, Operand<TFloat32> inputMin, Operand<TFloat32> inputMax, Class<U> outType) Factory method to create a class wrapping a new QuantizeDownAndShrinkRange operation.static RequantizationRangeRequantizationRange.create(Scope scope, Operand<? extends TNumber> input, Operand<TFloat32> inputMin, Operand<TFloat32> inputMax) Factory method to create a class wrapping a new RequantizationRange operation.static <U extends TNumber>
Requantize<U> Requantize.create(Scope scope, Operand<? extends TNumber> input, Operand<TFloat32> inputMin, Operand<TFloat32> inputMax, Operand<TFloat32> requestedOutputMin, Operand<TFloat32> requestedOutputMax, Class<U> outType) Factory method to create a class wrapping a new Requantize operation.static <U extends TNumber>
UniformDequantize<U> UniformDequantize.create(Scope scope, Operand<? extends TNumber> input, Operand<TFloat32> scales, Operand<TInt32> zeroPoints, Class<U> Tout, Long quantizationMinVal, Long quantizationMaxVal, UniformDequantize.Options... options) Factory method to create a class wrapping a new UniformDequantize operation.static <U extends TNumber>
UniformQuantize<U> UniformQuantize.create(Scope scope, Operand<? extends TNumber> input, Operand<TFloat32> scales, Operand<TInt32> zeroPoints, Class<U> Tout, Long quantizationMinVal, Long quantizationMaxVal, UniformQuantize.Options... options) Factory method to create a class wrapping a new UniformQuantize operation.static <U extends TNumber, T extends TNumber>
UniformQuantizedDot<U> UniformQuantizedDot.create(Scope scope, Operand<T> lhs, Operand<T> rhs, Operand<TFloat32> lhsScales, Operand<TInt32> lhsZeroPoints, Operand<TFloat32> rhsScales, Operand<TInt32> rhsZeroPoints, Operand<TFloat32> outputScales, Operand<TInt32> outputZeroPoints, Class<U> Tout, Long lhsQuantizationMinVal, Long lhsQuantizationMaxVal, Long rhsQuantizationMinVal, Long rhsQuantizationMaxVal, Long outputQuantizationMinVal, Long outputQuantizationMaxVal, UniformQuantizedDot.Options... options) Factory method to create a class wrapping a new UniformQuantizedDot operation.static <V extends TNumber>
UniformQuantizedDotHybrid<V> UniformQuantizedDotHybrid.create(Scope scope, Operand<? extends TNumber> lhs, Operand<? extends TNumber> rhs, Operand<TFloat32> rhsScales, Operand<TInt32> rhsZeroPoints, Class<V> Tout, Long rhsQuantizationMinVal, Long rhsQuantizationMaxVal, UniformQuantizedDotHybrid.Options... options) Factory method to create a class wrapping a new UniformQuantizedDotHybrid operation.static <U extends TNumber>
UniformRequantize<U> UniformRequantize.create(Scope scope, Operand<? extends TNumber> input, Operand<TFloat32> inputScales, Operand<TInt32> inputZeroPoints, Operand<TFloat32> outputScales, Operand<TInt32> outputZeroPoints, Class<U> Tout, Long inputQuantizationMinVal, Long inputQuantizationMaxVal, Long outputQuantizationMinVal, Long outputQuantizationMaxVal, UniformRequantize.Options... options) Factory method to create a class wrapping a new UniformRequantize operation.Method parameters in org.tensorflow.op.quantization with type arguments of type OperandModifier and TypeMethodDescriptionstatic <T extends TType>
QuantizedConcat<T> QuantizedConcat.create(Scope scope, Operand<TInt32> concatDim, Iterable<Operand<T>> values, Iterable<Operand<TFloat32>> inputMins, Iterable<Operand<TFloat32>> inputMaxes) Factory method to create a class wrapping a new QuantizedConcat operation. -
Uses of Operand in org.tensorflow.op.ragged
Classes in org.tensorflow.op.ragged that implement OperandModifier and TypeClassDescriptionfinal classRaggedBincount<U extends TNumber>Counts the number of occurrences of each value in an integer array.final classRaggedTensorToTensor<U extends TType>Create a dense tensor from a ragged tensor, possibly altering its shape.final classEncodes aRaggedTensorinto avariantTensor.final classRaggedTensorToVariantGradient<U extends TType>Helper used to compute the gradient forRaggedTensorToVariant.Fields in org.tensorflow.op.ragged declared as OperandModifier and TypeFieldDescriptionRaggedFillEmptyRows.Inputs.defaultValueThe defaultValue inputRaggedTensorToTensor.Inputs.defaultValueThe default_value when the shape is larger than the ragged tensor.RaggedRange.Inputs.deltasThe deltas of each range.RaggedTensorToVariantGradient.Inputs.denseValuesShapeShape of the dense_values that was used as an input to the RaggedTensorToVariant op.RaggedTensorFromVariant.Inputs.encodedRaggedAvariantTensor containing encodedRaggedTensors.RaggedTensorToVariantGradient.Inputs.encodedRaggedGradAvariantTensor containing encodedRaggedTensorgradients.RaggedFillEmptyRowsGrad.Inputs.gradValuesThe gradValues inputRaggedGather.Inputs.indicesIndices in the outermost dimension ofparamsof the values that should be gathered.RaggedRange.Inputs.limitsThe limits of each range.RaggedFillEmptyRows.Inputs.nrowsThe nrows inputRaggedGather.Inputs.paramsDenseValuesTheflat_valuesfor theparamsRaggedTensor.RaggedFillEmptyRowsGrad.Inputs.reverseIndexMapThe reverseIndexMap inputRaggedTensorToVariantGradient.Inputs.rowSplitsOutermost row-splits that were used as input to the RaggedTensorToVariant op.RaggedTensorToSparse.Inputs.rtDenseValuesTheflat_valuesfor theRaggedTensor.RaggedTensorToVariant.Inputs.rtDenseValuesA Tensor representing the values of the inputRaggedTensor.RaggedTensorToTensor.Inputs.shapeThe desired shape of the output tensor.RaggedBincount.Inputs.sizeOutputnon-negative int scalarTensor.RaggedBincount.Inputs.splits1D int64Tensor.RaggedCountSparseOutput.Inputs.splitsTensor containing the row splits of the ragged tensor to count.RaggedRange.Inputs.startsThe starts of each range.RaggedFillEmptyRows.Inputs.valueRowidsThe valueRowids inputRaggedBincount.Inputs.values2D intTensor.RaggedCountSparseOutput.Inputs.valuesTensor containing values of the sparse tensor to count.RaggedFillEmptyRows.Inputs.valuesThe values inputRaggedTensorToTensor.Inputs.valuesA 1D tensor representing the values of the ragged tensor.RaggedBincount.Inputs.weightsis an int32, int64, float32, or float64Tensorwith the same shape asinput, or a length-0Tensor, in which case it acts as all weights equal to 1.RaggedCountSparseOutput.Inputs.weightsA Tensor of the same shape as indices containing per-index weight values.Fields in org.tensorflow.op.ragged with type parameters of type OperandModifier and TypeFieldDescriptionRaggedCross.Inputs.denseInputsThe tf.Tensor inputs.RaggedGather.Inputs.paramsNestedSplitsThenested_row_splitstensors that define the row-partitioning for theparamsRaggedTensor input.RaggedCross.Inputs.raggedRowSplitsThe row_splits tensor for each RaggedTensor input.RaggedCross.Inputs.raggedValuesThe values tensor for each RaggedTensor input.RaggedTensorToTensor.Inputs.rowPartitionTensorsThe rowPartitionTensors inputRaggedTensorToSparse.Inputs.rtNestedSplitsTherow_splitsfor theRaggedTensor.RaggedTensorToVariant.Inputs.rtNestedSplitsA list of one or more Tensors representing the splits of the inputRaggedTensor.RaggedCross.Inputs.sparseIndicesThe indices tensor for each SparseTensor input.RaggedCross.Inputs.sparseShapeThe dense_shape tensor for each SparseTensor input.RaggedCross.Inputs.sparseValuesThe values tensor for each SparseTensor input.Methods in org.tensorflow.op.ragged with parameters of type OperandModifier and TypeMethodDescriptionstatic <U extends TNumber, T extends TNumber>
RaggedBincount<U> RaggedBincount.create(Scope scope, Operand<TInt64> splits, Operand<T> values, Operand<T> sizeOutput, Operand<U> weights, RaggedBincount.Options... options) Factory method to create a class wrapping a new RaggedBincount operation.static <U extends TNumber>
RaggedCountSparseOutput<U> RaggedCountSparseOutput.create(Scope scope, Operand<TInt64> splits, Operand<? extends TNumber> values, Operand<U> weights, Boolean binaryOutput, RaggedCountSparseOutput.Options... options) Factory method to create a class wrapping a new RaggedCountSparseOutput operation.static <T extends TType>
RaggedFillEmptyRows<T> RaggedFillEmptyRows.create(Scope scope, Operand<TInt64> valueRowids, Operand<T> values, Operand<TInt64> nrows, Operand<T> defaultValue) Factory method to create a class wrapping a new RaggedFillEmptyRows operation.static <T extends TType>
RaggedFillEmptyRowsGrad<T> Factory method to create a class wrapping a new RaggedFillEmptyRowsGrad operation.static <T extends TNumber, U extends TType>
RaggedGather<T, U> RaggedGather.create(Scope scope, Iterable<Operand<T>> paramsNestedSplits, Operand<U> paramsDenseValues, Operand<? extends TNumber> indices, Long OUTPUTRAGGEDRANK) Factory method to create a class wrapping a new RaggedGather operation.static <T extends TNumber>
RaggedRange<TInt64, T> Factory method to create a class wrapping a new RaggedRange operation, with the default output types.static <U extends TNumber, T extends TNumber>
RaggedRange<U, T> RaggedRange.create(Scope scope, Operand<T> starts, Operand<T> limits, Operand<T> deltas, Class<U> Tsplits) Factory method to create a class wrapping a new RaggedRange operation.static <U extends TType>
RaggedTensorFromVariant<TInt64, U> RaggedTensorFromVariant.create(Scope scope, Operand<? extends TType> encodedRagged, Long inputRaggedRank, Long outputRaggedRank, Class<U> Tvalues) Factory method to create a class wrapping a new RaggedTensorFromVariant operation, with the default output types.static <T extends TNumber, U extends TType>
RaggedTensorFromVariant<T, U> RaggedTensorFromVariant.create(Scope scope, Operand<? extends TType> encodedRagged, Long inputRaggedRank, Long outputRaggedRank, Class<U> Tvalues, Class<T> Tsplits) Factory method to create a class wrapping a new RaggedTensorFromVariant operation.static <U extends TType>
RaggedTensorToSparse<U> RaggedTensorToSparse.create(Scope scope, Iterable<Operand<? extends TNumber>> rtNestedSplits, Operand<U> rtDenseValues) Factory method to create a class wrapping a new RaggedTensorToSparse operation.static <U extends TType>
RaggedTensorToTensor<U> RaggedTensorToTensor.create(Scope scope, Operand<? extends TNumber> shape, Operand<U> values, Operand<U> defaultValue, Iterable<Operand<? extends TNumber>> rowPartitionTensors, List<String> rowPartitionTypes) Factory method to create a class wrapping a new RaggedTensorToTensor operation.static RaggedTensorToVariantRaggedTensorToVariant.create(Scope scope, Iterable<Operand<? extends TNumber>> rtNestedSplits, Operand<? extends TType> rtDenseValues, Boolean batchedInput) Factory method to create a class wrapping a new RaggedTensorToVariant operation.static <U extends TType>
RaggedTensorToVariantGradient<U> RaggedTensorToVariantGradient.create(Scope scope, Operand<? extends TType> encodedRaggedGrad, Operand<? extends TNumber> rowSplits, Operand<TInt32> denseValuesShape, Class<U> Tvalues) Factory method to create a class wrapping a new RaggedTensorToVariantGradient operation.Method parameters in org.tensorflow.op.ragged with type arguments of type OperandModifier and TypeMethodDescriptionstatic <T extends TType, U extends TNumber>
RaggedCross<T, U> RaggedCross.create(Scope scope, Iterable<Operand<?>> raggedValues, Iterable<Operand<?>> raggedRowSplits, Iterable<Operand<TInt64>> sparseIndices, Iterable<Operand<?>> sparseValues, Iterable<Operand<TInt64>> sparseShape, Iterable<Operand<?>> denseInputs, String inputOrder, Boolean hashedOutput, Long numBuckets, Long hashKey, Class<T> outValuesType, Class<U> outRowSplitsType) Factory method to create a class wrapping a new RaggedCross operation.static <T extends TNumber, U extends TType>
RaggedGather<T, U> RaggedGather.create(Scope scope, Iterable<Operand<T>> paramsNestedSplits, Operand<U> paramsDenseValues, Operand<? extends TNumber> indices, Long OUTPUTRAGGEDRANK) Factory method to create a class wrapping a new RaggedGather operation.static <U extends TType>
RaggedTensorToSparse<U> RaggedTensorToSparse.create(Scope scope, Iterable<Operand<? extends TNumber>> rtNestedSplits, Operand<U> rtDenseValues) Factory method to create a class wrapping a new RaggedTensorToSparse operation.static <U extends TType>
RaggedTensorToTensor<U> RaggedTensorToTensor.create(Scope scope, Operand<? extends TNumber> shape, Operand<U> values, Operand<U> defaultValue, Iterable<Operand<? extends TNumber>> rowPartitionTensors, List<String> rowPartitionTypes) Factory method to create a class wrapping a new RaggedTensorToTensor operation.static RaggedTensorToVariantRaggedTensorToVariant.create(Scope scope, Iterable<Operand<? extends TNumber>> rtNestedSplits, Operand<? extends TType> rtDenseValues, Boolean batchedInput) Factory method to create a class wrapping a new RaggedTensorToVariant operation. -
Uses of Operand in org.tensorflow.op.random
Classes in org.tensorflow.op.random that implement OperandModifier and TypeClassDescriptionfinal classThe DummySeedGenerator operationfinal classMultinomial<U extends TNumber>Draws samples from a multinomial distribution.final classNonDeterministicInts<U extends TType>Non-deterministically generates some integers.final classParameterizedTruncatedNormal<U extends TNumber>Outputs random values from a normal distribution.final classRandomGamma<U extends TNumber>Outputs random values from the Gamma distribution(s) described by alpha.final classRandomGammaGrad<T extends TNumber>Computes the derivative of a Gamma random sample w.r.t.final classRandomPoisson<V extends TNumber>Outputs random values from the Poisson distribution(s) described by rate.final classRandomShuffle<T extends TType>Randomly shuffles a tensor along its first dimension.final classRandomStandardNormal<U extends TNumber>Outputs random values from a normal distribution.final classRandomUniform<U extends TNumber>Outputs random values from a uniform distribution.final classRandomUniformInt<U extends TNumber>Outputs random integers from a uniform distribution.final classEmits randomized records.final classAdvance the counter of a counter-based RNG.final classStatefulRandomBinomial<V extends TNumber>The StatefulRandomBinomial operationfinal classStatefulStandardNormal<U extends TType>Outputs random values from a normal distribution.final classStatefulTruncatedNormal<U extends TType>Outputs random values from a truncated normal distribution.final classStatefulUniform<U extends TType>Outputs random values from a uniform distribution.final classStatefulUniformFullInt<U extends TType>Outputs random integers from a uniform distribution.final classStatefulUniformInt<U extends TType>Outputs random integers from a uniform distribution.final classStatelessMultinomial<V extends TNumber>Draws samples from a multinomial distribution.final classStatelessParameterizedTruncatedNormal<V extends TNumber>The StatelessParameterizedTruncatedNormal operationfinal classStatelessRandomBinomial<W extends TNumber>Outputs deterministic pseudorandom random numbers from a binomial distribution.final classStatelessRandomGamma<U extends TNumber>Outputs deterministic pseudorandom random numbers from a gamma distribution.final classPicks the best counter-based RNG algorithm based on device.final classStatelessRandomNormal<V extends TNumber>Outputs deterministic pseudorandom values from a normal distribution.final classStatelessRandomNormalV2<U extends TNumber>Outputs deterministic pseudorandom values from a normal distribution.final classStatelessRandomPoisson<W extends TNumber>Outputs deterministic pseudorandom random numbers from a Poisson distribution.final classStatelessRandomUniform<V extends TNumber>Outputs deterministic pseudorandom random values from a uniform distribution.final classStatelessRandomUniformFullInt<V extends TNumber>Outputs deterministic pseudorandom random integers from a uniform distribution.final classStatelessRandomUniformFullIntV2<U extends TNumber>Outputs deterministic pseudorandom random integers from a uniform distribution.final classStatelessRandomUniformInt<V extends TNumber>Outputs deterministic pseudorandom random integers from a uniform distribution.final classStatelessRandomUniformIntV2<U extends TNumber>Outputs deterministic pseudorandom random integers from a uniform distribution.final classStatelessRandomUniformV2<U extends TNumber>Outputs deterministic pseudorandom random values from a uniform distribution.final classStatelessTruncatedNormal<V extends TNumber>Outputs deterministic pseudorandom values from a truncated normal distribution.final classStatelessTruncatedNormalV2<U extends TNumber>Outputs deterministic pseudorandom values from a truncated normal distribution.final classTruncatedNormal<U extends TNumber>Outputs random values from a truncated normal distribution.Fields in org.tensorflow.op.random declared as OperandModifier and TypeFieldDescriptionRngReadAndSkip.Inputs.algThe RNG algorithm.StatelessRandomGamma.Inputs.algThe RNG algorithm (shape int32[]).StatelessRandomNormalV2.Inputs.algThe RNG algorithm (shape int32[]).StatelessRandomUniformFullIntV2.Inputs.algThe RNG algorithm (shape int32[]).StatelessRandomUniformIntV2.Inputs.algThe RNG algorithm (shape int32[]).StatelessRandomUniformV2.Inputs.algThe RNG algorithm (shape int32[]).StatelessTruncatedNormalV2.Inputs.algThe RNG algorithm (shape int32[]).RngSkip.Inputs.algorithmThe RNG algorithm.StatefulRandomBinomial.Inputs.algorithmThe algorithm inputStatefulStandardNormal.Inputs.algorithmThe RNG algorithm.StatefulTruncatedNormal.Inputs.algorithmThe RNG algorithm.StatefulUniform.Inputs.algorithmThe RNG algorithm.StatefulUniformFullInt.Inputs.algorithmThe RNG algorithm.StatefulUniformInt.Inputs.algorithmThe RNG algorithm.RandomGamma.Inputs.alphaA tensor in which each scalar is a "shape" parameter describing the associated gamma distribution.RandomGammaGrad.Inputs.alphaThe alpha inputStatelessRandomGamma.Inputs.alphaThe concentration of the gamma distribution.StatelessRandomGamma.Inputs.counterInitial counter for the counter-based RNG algorithm (shape uint64[2] or uint64[1] depending on the algorithm).StatelessRandomNormalV2.Inputs.counterInitial counter for the counter-based RNG algorithm (shape uint64[2] or uint64[1] depending on the algorithm).StatelessRandomUniformFullIntV2.Inputs.counterInitial counter for the counter-based RNG algorithm (shape uint64[2] or uint64[1] depending on the algorithm).StatelessRandomUniformIntV2.Inputs.counterInitial counter for the counter-based RNG algorithm (shape uint64[2] or uint64[1] depending on the algorithm).StatelessRandomUniformV2.Inputs.counterInitial counter for the counter-based RNG algorithm (shape uint64[2] or uint64[1] depending on the algorithm).StatelessTruncatedNormalV2.Inputs.counterInitial counter for the counter-based RNG algorithm (shape uint64[2] or uint64[1] depending on the algorithm).StatefulRandomBinomial.Inputs.countsThe counts inputStatelessRandomBinomial.Inputs.countsThe counts of the binomial distribution.DeleteRandomSeedGenerator.Inputs.deleterThe deleter inputDeleteSeedGenerator.Inputs.deleterThe deleter inputRngReadAndSkip.Inputs.deltaThe amount of advancement.RngSkip.Inputs.deltaThe amount of advancement.DeleteRandomSeedGenerator.Inputs.handleThe handle inputDeleteSeedGenerator.Inputs.handleThe handle inputStatelessRandomGamma.Inputs.keyKey for the counter-based RNG algorithm (shape uint64[1]).StatelessRandomNormalV2.Inputs.keyKey for the counter-based RNG algorithm (shape uint64[1]).StatelessRandomUniformFullIntV2.Inputs.keyKey for the counter-based RNG algorithm (shape uint64[1]).StatelessRandomUniformIntV2.Inputs.keyKey for the counter-based RNG algorithm (shape uint64[1]).StatelessRandomUniformV2.Inputs.keyKey for the counter-based RNG algorithm (shape uint64[1]).StatelessTruncatedNormalV2.Inputs.keyKey for the counter-based RNG algorithm (shape uint64[1]).StatelessRandomPoisson.Inputs.lamThe rate of the Poisson distribution.Multinomial.Inputs.logits2-D Tensor with shape[batch_size, num_classes].StatelessMultinomial.Inputs.logits2-D Tensor with shape[batch_size, num_classes].RandomUniformInt.Inputs.maxval0-D.StatefulUniformInt.Inputs.maxvalMaximum value (exclusive, scalar).StatelessRandomUniformInt.Inputs.maxvalMaximum value (exclusive, scalar).StatelessRandomUniformIntV2.Inputs.maxvalMaximum value (exclusive, scalar).ParameterizedTruncatedNormal.Inputs.maxvalsThe maximum cutoff.StatelessParameterizedTruncatedNormal.Inputs.maxvalsThe maximum cutoff.ParameterizedTruncatedNormal.Inputs.meansThe mean parameter of each batch.StatelessParameterizedTruncatedNormal.Inputs.meansThe mean parameter of each batch.RandomUniformInt.Inputs.minval0-D.StatefulUniformInt.Inputs.minvalMinimum value (inclusive, scalar).StatelessRandomUniformInt.Inputs.minvalMinimum value (inclusive, scalar).StatelessRandomUniformIntV2.Inputs.minvalMinimum value (inclusive, scalar).ParameterizedTruncatedNormal.Inputs.minvalsThe minimum cutoff.StatelessParameterizedTruncatedNormal.Inputs.minvalsThe minimum cutoff.Multinomial.Inputs.numSamples0-D.StatelessMultinomial.Inputs.numSamples0-D.StatefulRandomBinomial.Inputs.probsThe probs inputStatelessRandomBinomial.Inputs.probsThe probability of success for the binomial distribution.RandomPoisson.Inputs.rateA tensor in which each scalar is a "rate" parameter describing the associated poisson distribution.AnonymousSeedGenerator.Inputs.reshuffleThe reshuffle inputRngReadAndSkip.Inputs.resourceThe handle of the resource variable that stores the state of the RNG.RngSkip.Inputs.resourceThe handle of the resource variable that stores the state of the RNG.StatefulRandomBinomial.Inputs.resourceThe resource inputStatefulStandardNormal.Inputs.resourceThe handle of the resource variable that stores the state of the RNG.StatefulTruncatedNormal.Inputs.resourceThe handle of the resource variable that stores the state of the RNG.StatefulUniform.Inputs.resourceThe handle of the resource variable that stores the state of the RNG.StatefulUniformFullInt.Inputs.resourceThe handle of the resource variable that stores the state of the RNG.StatefulUniformInt.Inputs.resourceThe handle of the resource variable that stores the state of the RNG.RandomGammaGrad.Inputs.sampleThe sample inputAnonymousRandomSeedGenerator.Inputs.seedThe seed inputAnonymousSeedGenerator.Inputs.seedThe seed inputStatelessMultinomial.Inputs.seed2 seeds (shape [2]).StatelessParameterizedTruncatedNormal.Inputs.seed2 seeds (shape [2]).StatelessRandomBinomial.Inputs.seed2 seeds (shape [2]).StatelessRandomGetKeyCounter.Inputs.seed2 seeds (shape [2]).StatelessRandomGetKeyCounterAlg.Inputs.seed2 seeds (shape [2]).StatelessRandomNormal.Inputs.seed2 seeds (shape [2]).StatelessRandomPoisson.Inputs.seed2 seeds (shape [2]).StatelessRandomUniform.Inputs.seed2 seeds (shape [2]).StatelessRandomUniformFullInt.Inputs.seed2 seeds (shape [2]).StatelessRandomUniformInt.Inputs.seed2 seeds (shape [2]).StatelessTruncatedNormal.Inputs.seed2 seeds (shape [2]).AnonymousRandomSeedGenerator.Inputs.seed2The seed2 inputAnonymousSeedGenerator.Inputs.seed2The seed2 inputNonDeterministicInts.Inputs.shapeThe shape of the output tensor.ParameterizedTruncatedNormal.Inputs.shapeThe shape of the output tensor.RandomGamma.Inputs.shape1-D integer tensor.RandomPoisson.Inputs.shape1-D integer tensor.RandomStandardNormal.Inputs.shapeThe shape of the output tensor.RandomUniform.Inputs.shapeThe shape of the output tensor.RandomUniformInt.Inputs.shapeThe shape of the output tensor.StatefulRandomBinomial.Inputs.shapeThe shape inputStatefulStandardNormal.Inputs.shapeThe shape of the output tensor.StatefulTruncatedNormal.Inputs.shapeThe shape of the output tensor.StatefulUniform.Inputs.shapeThe shape of the output tensor.StatefulUniformFullInt.Inputs.shapeThe shape of the output tensor.StatefulUniformInt.Inputs.shapeThe shape of the output tensor.StatelessParameterizedTruncatedNormal.Inputs.shapeThe shape of the output tensor.StatelessRandomBinomial.Inputs.shapeThe shape of the output tensor.StatelessRandomGamma.Inputs.shapeThe shape of the output tensor.StatelessRandomNormal.Inputs.shapeThe shape of the output tensor.StatelessRandomNormalV2.Inputs.shapeThe shape of the output tensor.StatelessRandomPoisson.Inputs.shapeThe shape of the output tensor.StatelessRandomUniform.Inputs.shapeThe shape of the output tensor.StatelessRandomUniformFullInt.Inputs.shapeThe shape of the output tensor.StatelessRandomUniformFullIntV2.Inputs.shapeThe shape of the output tensor.StatelessRandomUniformInt.Inputs.shapeThe shape of the output tensor.StatelessRandomUniformIntV2.Inputs.shapeThe shape of the output tensor.StatelessRandomUniformV2.Inputs.shapeThe shape of the output tensor.StatelessTruncatedNormal.Inputs.shapeThe shape of the output tensor.StatelessTruncatedNormalV2.Inputs.shapeThe shape of the output tensor.TruncatedNormal.Inputs.shapeThe shape of the output tensor.StatelessParameterizedTruncatedNormal.Inputs.stddevsThe standard deviation parameter of each batch.ParameterizedTruncatedNormal.Inputs.stdevsThe standard deviation parameter of each batch.AllCandidateSampler.Inputs.trueClassesA batch_size * num_true matrix, in which each row contains the IDs of the num_true target_classes in the corresponding original label.LogUniformCandidateSampler.Inputs.trueClassesA batch_size * num_true matrix, in which each row contains the IDs of the num_true target_classes in the corresponding original label.ThreadUnsafeUnigramCandidateSampler.Inputs.trueClassesA batch_size * num_true matrix, in which each row contains the IDs of the num_true target_classes in the corresponding original label.UniformCandidateSampler.Inputs.trueClassesA batch_size * num_true matrix, in which each row contains the IDs of the num_true target_classes in the corresponding original label.RandomShuffle.Inputs.valueThe tensor to be shuffled.Methods in org.tensorflow.op.random with parameters of type OperandModifier and TypeMethodDescriptionstatic AllCandidateSamplerAllCandidateSampler.create(Scope scope, Operand<TInt64> trueClasses, Long numTrue, Long numSampled, Boolean unique, AllCandidateSampler.Options... options) Factory method to create a class wrapping a new AllCandidateSampler operation.static AnonymousRandomSeedGeneratorFactory method to create a class wrapping a new AnonymousRandomSeedGenerator operation.static AnonymousSeedGeneratorAnonymousSeedGenerator.create(Scope scope, Operand<TInt64> seed, Operand<TInt64> seed2, Operand<TBool> reshuffle) Factory method to create a class wrapping a new AnonymousSeedGenerator operation.static DeleteRandomSeedGeneratorDeleteRandomSeedGenerator.create(Scope scope, Operand<? extends TType> handle, Operand<? extends TType> deleter) Factory method to create a class wrapping a new DeleteRandomSeedGenerator operation.static DeleteSeedGeneratorDeleteSeedGenerator.create(Scope scope, Operand<? extends TType> handle, Operand<? extends TType> deleter) Factory method to create a class wrapping a new DeleteSeedGenerator operation.static LogUniformCandidateSamplerLogUniformCandidateSampler.create(Scope scope, Operand<TInt64> trueClasses, Long numTrue, Long numSampled, Boolean unique, Long rangeMax, LogUniformCandidateSampler.Options... options) Factory method to create a class wrapping a new LogUniformCandidateSampler operation.static <U extends TNumber>
Multinomial<U> Multinomial.create(Scope scope, Operand<? extends TNumber> logits, Operand<TInt32> numSamples, Class<U> outputDtype, Multinomial.Options... options) Factory method to create a class wrapping a new Multinomial operation.static Multinomial<TInt64> Multinomial.create(Scope scope, Operand<? extends TNumber> logits, Operand<TInt32> numSamples, Multinomial.Options[] options) Factory method to create a class wrapping a new Multinomial operation, with the default output types.static NonDeterministicInts<TInt64> Factory method to create a class wrapping a new NonDeterministicInts operation, with the default output types.static <U extends TType>
NonDeterministicInts<U> Factory method to create a class wrapping a new NonDeterministicInts operation.static <U extends TNumber>
ParameterizedTruncatedNormal<U> ParameterizedTruncatedNormal.create(Scope scope, Operand<? extends TNumber> shape, Operand<U> means, Operand<U> stdevs, Operand<U> minvals, Operand<U> maxvals, ParameterizedTruncatedNormal.Options... options) Factory method to create a class wrapping a new ParameterizedTruncatedNormal operation.static <U extends TNumber>
RandomGamma<U> RandomGamma.create(Scope scope, Operand<? extends TNumber> shape, Operand<U> alpha, RandomGamma.Options... options) Factory method to create a class wrapping a new RandomGamma operation.static <T extends TNumber>
RandomGammaGrad<T> Factory method to create a class wrapping a new RandomGammaGrad operation.static <V extends TNumber>
RandomPoisson<V> RandomPoisson.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TNumber> rate, Class<V> dtype, RandomPoisson.Options... options) Factory method to create a class wrapping a new RandomPoissonV2 operation.static RandomPoisson<TInt64> RandomPoisson.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TNumber> rate, RandomPoisson.Options[] options) Factory method to create a class wrapping a new RandomPoissonV2 operation, with the default output types.static <T extends TType>
RandomShuffle<T> RandomShuffle.create(Scope scope, Operand<T> value, RandomShuffle.Options... options) Factory method to create a class wrapping a new RandomShuffle operation.static <U extends TNumber>
RandomStandardNormal<U> RandomStandardNormal.create(Scope scope, Operand<? extends TNumber> shape, Class<U> dtype, RandomStandardNormal.Options... options) Factory method to create a class wrapping a new RandomStandardNormal operation.static <U extends TNumber>
RandomUniform<U> RandomUniform.create(Scope scope, Operand<? extends TNumber> shape, Class<U> dtype, RandomUniform.Options... options) Factory method to create a class wrapping a new RandomUniform operation.static <U extends TNumber>
RandomUniformInt<U> RandomUniformInt.create(Scope scope, Operand<? extends TNumber> shape, Operand<U> minval, Operand<U> maxval, RandomUniformInt.Options... options) Factory method to create a class wrapping a new RandomUniformInt operation.static RngReadAndSkipRngReadAndSkip.create(Scope scope, Operand<? extends TType> resource, Operand<TInt32> alg, Operand<? extends TType> delta) Factory method to create a class wrapping a new RngReadAndSkip operation.static RngSkipRngSkip.create(Scope scope, Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<TInt64> delta) Factory method to create a class wrapping a new RngSkip operation.static <U extends TNumber>
StatefulRandomBinomial<TInt64> StatefulRandomBinomial.create(Scope scope, Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TNumber> shape, Operand<U> counts, Operand<U> probs) Factory method to create a class wrapping a new StatefulRandomBinomial operation, with the default output types.static <V extends TNumber, U extends TNumber>
StatefulRandomBinomial<V> StatefulRandomBinomial.create(Scope scope, Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TNumber> shape, Operand<U> counts, Operand<U> probs, Class<V> dtype) Factory method to create a class wrapping a new StatefulRandomBinomial operation.static StatefulStandardNormal<TFloat32> StatefulStandardNormal.create(Scope scope, Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TType> shape) Factory method to create a class wrapping a new StatefulStandardNormalV2 operation, with the default output types.static <U extends TType>
StatefulStandardNormal<U> StatefulStandardNormal.create(Scope scope, Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TType> shape, Class<U> dtype) Factory method to create a class wrapping a new StatefulStandardNormalV2 operation.static StatefulTruncatedNormal<TFloat32> StatefulTruncatedNormal.create(Scope scope, Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TType> shape) Factory method to create a class wrapping a new StatefulTruncatedNormal operation, with the default output types.static <U extends TType>
StatefulTruncatedNormal<U> StatefulTruncatedNormal.create(Scope scope, Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TType> shape, Class<U> dtype) Factory method to create a class wrapping a new StatefulTruncatedNormal operation.static StatefulUniform<TFloat32> StatefulUniform.create(Scope scope, Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TType> shape) Factory method to create a class wrapping a new StatefulUniform operation, with the default output types.static <U extends TType>
StatefulUniform<U> StatefulUniform.create(Scope scope, Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TType> shape, Class<U> dtype) Factory method to create a class wrapping a new StatefulUniform operation.static <U extends TType>
StatefulUniformFullInt<U> StatefulUniformFullInt.create(Scope scope, Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TType> shape, Class<U> dtype) Factory method to create a class wrapping a new StatefulUniformFullInt operation.static <U extends TType>
StatefulUniformInt<U> StatefulUniformInt.create(Scope scope, Operand<? extends TType> resource, Operand<TInt64> algorithm, Operand<? extends TType> shape, Operand<U> minval, Operand<U> maxval) Factory method to create a class wrapping a new StatefulUniformInt operation.static StatelessMultinomial<TInt64> StatelessMultinomial.create(Scope scope, Operand<? extends TNumber> logits, Operand<TInt32> numSamples, Operand<? extends TNumber> seed) Factory method to create a class wrapping a new StatelessMultinomial operation, with the default output types.static <V extends TNumber>
StatelessMultinomial<V> StatelessMultinomial.create(Scope scope, Operand<? extends TNumber> logits, Operand<TInt32> numSamples, Operand<? extends TNumber> seed, Class<V> outputDtype) Factory method to create a class wrapping a new StatelessMultinomial operation.static <V extends TNumber>
StatelessParameterizedTruncatedNormal<V> StatelessParameterizedTruncatedNormal.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TNumber> seed, Operand<V> means, Operand<V> stddevs, Operand<V> minvals, Operand<V> maxvals) Factory method to create a class wrapping a new StatelessParameterizedTruncatedNormal operation.static <V extends TNumber>
StatelessRandomBinomial<TInt64> StatelessRandomBinomial.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TNumber> seed, Operand<V> counts, Operand<V> probs) Factory method to create a class wrapping a new StatelessRandomBinomial operation, with the default output types.static <W extends TNumber, V extends TNumber>
StatelessRandomBinomial<W> StatelessRandomBinomial.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TNumber> seed, Operand<V> counts, Operand<V> probs, Class<W> dtype) Factory method to create a class wrapping a new StatelessRandomBinomial operation.static <U extends TNumber>
StatelessRandomGamma<U> StatelessRandomGamma.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg, Operand<U> alpha) Factory method to create a class wrapping a new StatelessRandomGammaV3 operation.static StatelessRandomGetKeyCounterFactory method to create a class wrapping a new StatelessRandomGetKeyCounter operation.Factory method to create a class wrapping a new StatelessRandomGetKeyCounterAlg operation.static StatelessRandomNormal<TFloat32> StatelessRandomNormal.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TNumber> seed) Factory method to create a class wrapping a new StatelessRandomNormal operation, with the default output types.static <V extends TNumber>
StatelessRandomNormal<V> StatelessRandomNormal.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TNumber> seed, Class<V> dtype) Factory method to create a class wrapping a new StatelessRandomNormal operation.static StatelessRandomNormalV2<TFloat32> StatelessRandomNormalV2.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg) Factory method to create a class wrapping a new StatelessRandomNormalV2 operation, with the default output types.static <U extends TNumber>
StatelessRandomNormalV2<U> StatelessRandomNormalV2.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg, Class<U> dtype) Factory method to create a class wrapping a new StatelessRandomNormalV2 operation.static <W extends TNumber>
StatelessRandomPoisson<W> StatelessRandomPoisson.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TNumber> seed, Operand<? extends TNumber> lam, Class<W> dtype) Factory method to create a class wrapping a new StatelessRandomPoisson operation.static StatelessRandomUniform<TFloat32> StatelessRandomUniform.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TNumber> seed) Factory method to create a class wrapping a new StatelessRandomUniform operation, with the default output types.static <V extends TNumber>
StatelessRandomUniform<V> StatelessRandomUniform.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TNumber> seed, Class<V> dtype) Factory method to create a class wrapping a new StatelessRandomUniform operation.static <V extends TNumber>
StatelessRandomUniformFullInt<V> StatelessRandomUniformFullInt.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TNumber> seed, Class<V> dtype) Factory method to create a class wrapping a new StatelessRandomUniformFullInt operation.static <U extends TNumber>
StatelessRandomUniformFullIntV2<U> StatelessRandomUniformFullIntV2.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg, Class<U> dtype) Factory method to create a class wrapping a new StatelessRandomUniformFullIntV2 operation.static <V extends TNumber>
StatelessRandomUniformInt<V> StatelessRandomUniformInt.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TNumber> seed, Operand<V> minval, Operand<V> maxval) Factory method to create a class wrapping a new StatelessRandomUniformInt operation.static <U extends TNumber>
StatelessRandomUniformIntV2<U> StatelessRandomUniformIntV2.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg, Operand<U> minval, Operand<U> maxval) Factory method to create a class wrapping a new StatelessRandomUniformIntV2 operation.static StatelessRandomUniformV2<TFloat32> StatelessRandomUniformV2.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg) Factory method to create a class wrapping a new StatelessRandomUniformV2 operation, with the default output types.static <U extends TNumber>
StatelessRandomUniformV2<U> StatelessRandomUniformV2.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg, Class<U> dtype) Factory method to create a class wrapping a new StatelessRandomUniformV2 operation.static StatelessTruncatedNormal<TFloat32> StatelessTruncatedNormal.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TNumber> seed) Factory method to create a class wrapping a new StatelessTruncatedNormal operation, with the default output types.static <V extends TNumber>
StatelessTruncatedNormal<V> StatelessTruncatedNormal.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TNumber> seed, Class<V> dtype) Factory method to create a class wrapping a new StatelessTruncatedNormal operation.static StatelessTruncatedNormalV2<TFloat32> StatelessTruncatedNormalV2.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg) Factory method to create a class wrapping a new StatelessTruncatedNormalV2 operation, with the default output types.static <U extends TNumber>
StatelessTruncatedNormalV2<U> StatelessTruncatedNormalV2.create(Scope scope, Operand<? extends TNumber> shape, Operand<? extends TType> key, Operand<? extends TType> counter, Operand<TInt32> alg, Class<U> dtype) Factory method to create a class wrapping a new StatelessTruncatedNormalV2 operation.ThreadUnsafeUnigramCandidateSampler.create(Scope scope, Operand<TInt64> trueClasses, Long numTrue, Long numSampled, Boolean unique, Long rangeMax, ThreadUnsafeUnigramCandidateSampler.Options... options) Factory method to create a class wrapping a new ThreadUnsafeUnigramCandidateSampler operation.static <U extends TNumber>
TruncatedNormal<U> TruncatedNormal.create(Scope scope, Operand<? extends TNumber> shape, Class<U> dtype, TruncatedNormal.Options... options) Factory method to create a class wrapping a new TruncatedNormal operation.static UniformCandidateSamplerUniformCandidateSampler.create(Scope scope, Operand<TInt64> trueClasses, Long numTrue, Long numSampled, Boolean unique, Long rangeMax, UniformCandidateSampler.Options... options) Factory method to create a class wrapping a new UniformCandidateSampler operation. -
Uses of Operand in org.tensorflow.op.random.experimental
Classes in org.tensorflow.op.random.experimental that implement OperandModifier and TypeClassDescriptionfinal classStatelessShuffle<T extends TType>Randomly and deterministically shuffles a tensor along its first dimension.Fields in org.tensorflow.op.random.experimental declared as OperandModifier and TypeFieldDescriptionStatelessShuffle.Inputs.algThe RNG algorithm (shape int32[]).StatelessShuffle.Inputs.counterInitial counter for the counter-based RNG algorithm (shape uint64[2] or uint64[1] depending on the algorithm).StatelessShuffle.Inputs.keyKey for the counter-based RNG algorithm (shape uint64[1]).StatelessShuffle.Inputs.valueThe tensor to be shuffled.Methods in org.tensorflow.op.random.experimental with parameters of type Operand -
Uses of Operand in org.tensorflow.op.signal
Classes in org.tensorflow.op.signal that implement OperandModifier and TypeClassDescriptionfinal classThe BatchFFT operationfinal classThe BatchFFT2D operationfinal classThe BatchFFT3D operationfinal classThe BatchIFFT operationfinal classThe BatchIFFT2D operationfinal classThe BatchIFFT3D operationfinal classFast Fourier transform.final class2D fast Fourier transform.final class3D fast Fourier transform.final classND fast Fourier transform.final classInverse fast Fourier transform.final classInverse 2D fast Fourier transform.final classInverse 3D fast Fourier transform.final classND inverse fast Fourier transform.final classInverse real-valued fast Fourier transform.final classInverse 2D real-valued fast Fourier transform.final classInverse 3D real-valued fast Fourier transform.final classND inverse real fast Fourier transform.final classReal-valued fast Fourier transform.final class2D real-valued fast Fourier transform.final class3D real-valued fast Fourier transform.final classND fast real Fourier transform.Fields in org.tensorflow.op.signal declared as OperandModifier and TypeFieldDescriptionFftNd.Inputs.axesAn int32 tensor with a same shape as fft_length.IfftNd.Inputs.axesAn int32 tensor with a same shape as fft_length.IrfftNd.Inputs.axesAn int32 tensor with a same shape as fft_length.RfftNd.Inputs.axesAn int32 tensor with a same shape as fft_length.FftNd.Inputs.fftLengthAn int32 tensor.IfftNd.Inputs.fftLengthAn int32 tensor.Irfft.Inputs.fftLengthAn int32 tensor of shape [1].Irfft2d.Inputs.fftLengthAn int32 tensor of shape [2].Irfft3d.Inputs.fftLengthAn int32 tensor of shape [3].IrfftNd.Inputs.fftLengthAn int32 tensor.Rfft.Inputs.fftLengthAn int32 tensor of shape [1].Rfft2d.Inputs.fftLengthAn int32 tensor of shape [2].Rfft3d.Inputs.fftLengthAn int32 tensor of shape [3].RfftNd.Inputs.fftLengthAn int32 tensor.BatchFft.Inputs.inputThe input inputBatchFft2d.Inputs.inputThe input inputBatchFft3d.Inputs.inputThe input inputBatchIfft.Inputs.inputThe input inputBatchIfft2d.Inputs.inputThe input inputBatchIfft3d.Inputs.inputThe input inputFft.Inputs.inputA complex tensor.Fft2d.Inputs.inputA complex tensor.Fft3d.Inputs.inputA complex tensor.FftNd.Inputs.inputA complex tensor.Ifft.Inputs.inputA complex tensor.Ifft2d.Inputs.inputA complex tensor.Ifft3d.Inputs.inputA complex tensor.IfftNd.Inputs.inputA complex tensor.Irfft.Inputs.inputA complex tensor.Irfft2d.Inputs.inputA complex tensor.Irfft3d.Inputs.inputA complex tensor.IrfftNd.Inputs.inputA complex tensor.Rfft.Inputs.inputA float32 tensor.Rfft2d.Inputs.inputA float32 tensor.Rfft3d.Inputs.inputA float32 tensor.RfftNd.Inputs.inputA complex tensor.Methods in org.tensorflow.op.signal with parameters of type OperandModifier and TypeMethodDescriptionstatic BatchFftFactory method to create a class wrapping a new BatchFFT operation.static BatchFft2dFactory method to create a class wrapping a new BatchFFT2D operation.static BatchFft3dFactory method to create a class wrapping a new BatchFFT3D operation.static BatchIfftFactory method to create a class wrapping a new BatchIFFT operation.static BatchIfft2dFactory method to create a class wrapping a new BatchIFFT2D operation.static BatchIfft3dFactory method to create a class wrapping a new BatchIFFT3D operation.Factory method to create a class wrapping a new FFT operation.Factory method to create a class wrapping a new FFT2D operation.Factory method to create a class wrapping a new FFT3D operation.Factory method to create a class wrapping a new FFTND operation.Factory method to create a class wrapping a new IFFT operation.Factory method to create a class wrapping a new IFFT2D operation.Factory method to create a class wrapping a new IFFT3D operation.Factory method to create a class wrapping a new IFFTND operation.Factory method to create a class wrapping a new IRFFT operation, with the default output types.Irfft.create(Scope scope, Operand<? extends TType> input, Operand<TInt32> fftLength, Class<U> Treal) Factory method to create a class wrapping a new IRFFT operation.Factory method to create a class wrapping a new IRFFT2D operation, with the default output types.Irfft2d.create(Scope scope, Operand<? extends TType> input, Operand<TInt32> fftLength, Class<U> Treal) Factory method to create a class wrapping a new IRFFT2D operation.Factory method to create a class wrapping a new IRFFT3D operation, with the default output types.Irfft3d.create(Scope scope, Operand<? extends TType> input, Operand<TInt32> fftLength, Class<U> Treal) Factory method to create a class wrapping a new IRFFT3D operation.IrfftNd.create(Scope scope, Operand<? extends TType> input, Operand<TInt32> fftLength, Operand<TInt32> axes) Factory method to create a class wrapping a new IRFFTND operation, with the default output types.IrfftNd.create(Scope scope, Operand<? extends TType> input, Operand<TInt32> fftLength, Operand<TInt32> axes, Class<U> Treal) Factory method to create a class wrapping a new IRFFTND operation.Rfft.create(Scope scope, Operand<? extends TNumber> input, Operand<TInt32> fftLength, Class<U> Tcomplex) Factory method to create a class wrapping a new RFFT operation.Rfft2d.create(Scope scope, Operand<? extends TNumber> input, Operand<TInt32> fftLength, Class<U> Tcomplex) Factory method to create a class wrapping a new RFFT2D operation.Rfft3d.create(Scope scope, Operand<? extends TNumber> input, Operand<TInt32> fftLength, Class<U> Tcomplex) Factory method to create a class wrapping a new RFFT3D operation.RfftNd.create(Scope scope, Operand<? extends TNumber> input, Operand<TInt32> fftLength, Operand<TInt32> axes, Class<U> Tcomplex) Factory method to create a class wrapping a new RFFTND operation. -
Uses of Operand in org.tensorflow.op.sparse
Classes in org.tensorflow.op.sparse that implement OperandModifier and TypeClassDescriptionfinal classAdd anN-minibatchSparseTensorto aSparseTensorsMap, returnNhandles.final classAdd aSparseTensorto aSparseTensorsMapreturn its handle.final classSparseBincount<U extends TNumber>Counts the number of occurrences of each value in an integer array.final classA conditional accumulator for aggregating sparse gradients.final classSparseDenseCwiseAdd<T extends TType>Adds up a SparseTensor and a dense Tensor, using these special rules: (1) Broadcasts the dense side to have the same shape as the sparse side, if eligible; (2) Then, only the dense values pointed to by the indices of the SparseTensor participate in the cwise addition.final classSparseDenseCwiseDiv<T extends TType>Component-wise divides a SparseTensor by a dense Tensor.final classSparseDenseCwiseMul<T extends TType>Component-wise multiplies a SparseTensor by a dense Tensor.final classMultiply matrix "a" by matrix "b".final classSparseReduceMax<T extends TNumber>Computes the max of elements across dimensions of a SparseTensor.final classSparseReduceSum<T extends TType>Computes the sum of elements across dimensions of a SparseTensor.final classSparseSegmentMean<T extends TNumber>Computes the mean along sparse segments of a tensor.final classSparseSegmentMeanWithNumSegments<T extends TNumber>Computes the mean along sparse segments of a tensor.final classSparseSegmentSqrtN<T extends TNumber>Computes the sum along sparse segments of a tensor divided by the sqrt of N.final classSparseSegmentSqrtNWithNumSegments<T extends TNumber>Computes the sum along sparse segments of a tensor divided by the sqrt of N.final classSparseSegmentSum<T extends TNumber>Computes the sum along sparse segments of a tensor.final classSparseSegmentSumWithNumSegments<T extends TNumber>Computes the sum along sparse segments of a tensor.final classSparseSliceGrad<T extends TType>The gradient operator for the SparseSlice op.final classSparseSoftmax<T extends TNumber>Applies softmax to a batched N-DSparseTensor.final classSparseTensorDenseAdd<U extends TType>Adds up aSparseTensorand a denseTensor, producing a denseTensor.final classSparseTensorDenseMatMul<U extends TType>Multiply SparseTensor (of rank 2) "A" by dense matrix "B".final classSparseToDense<U extends TType>Converts a sparse representation into a dense tensor.Fields in org.tensorflow.op.sparse declared as OperandModifier and TypeFieldDescriptionSparseMatMul.Inputs.aThe a inputSparseAdd.Inputs.aIndices2-D.SparseAddGrad.Inputs.aIndices2-D.SparseSparseMaximum.Inputs.aIndices2-D.SparseSparseMinimum.Inputs.aIndices2-D.SparseTensorDenseAdd.Inputs.aIndices2-D.SparseTensorDenseMatMul.Inputs.aIndices2-D.SparseAdd.Inputs.aShape1-D.SparseSparseMaximum.Inputs.aShape1-D.SparseSparseMinimum.Inputs.aShape1-D.SparseTensorDenseAdd.Inputs.aShape1-D.SparseTensorDenseMatMul.Inputs.aShape1-D.SparseAdd.Inputs.aValues1-D.SparseSparseMaximum.Inputs.aValues1-D.SparseSparseMinimum.Inputs.aValues1-D.SparseTensorDenseAdd.Inputs.aValues1-D.SparseTensorDenseMatMul.Inputs.aValues1-D.SparseMatMul.Inputs.bThe b inputSparseTensorDenseAdd.Inputs.bndims-D Tensor.SparseTensorDenseMatMul.Inputs.b2-D.SparseAddGrad.Inputs.backpropValGrad1-D with shape[nnz(sum)].SparseSliceGrad.Inputs.backpropValGrad1-D.SparseAdd.Inputs.bIndices2-D.SparseAddGrad.Inputs.bIndices2-D.SparseSparseMaximum.Inputs.bIndicescounterpart toa_indicesfor the other operand.SparseSparseMinimum.Inputs.bIndicescounterpart toa_indicesfor the other operand.SparseAdd.Inputs.bShape1-D.SparseSparseMaximum.Inputs.bShapecounterpart toa_shapefor the other operand; the two shapes must be equal.SparseSparseMinimum.Inputs.bShapecounterpart toa_shapefor the other operand; the two shapes must be equal.SparseAdd.Inputs.bValues1-D.SparseSparseMaximum.Inputs.bValuescounterpart toa_valuesfor the other operand; must be of the same dtype.SparseSparseMinimum.Inputs.bValuescounterpart toa_valuesfor the other operand; must be of the same dtype.SparseSegmentMean.Inputs.dataThe data inputSparseSegmentMeanWithNumSegments.Inputs.dataThe data inputSparseSegmentSqrtN.Inputs.dataThe data inputSparseSegmentSqrtNWithNumSegments.Inputs.dataThe data inputSparseSegmentSum.Inputs.dataThe data inputSparseSegmentSumWithNumSegments.Inputs.dataThe data inputSparseFillEmptyRows.Inputs.defaultValue0-D. default value to insert into location[row, 0, ..., 0]for rows missing from the input sparse tensor. output indices: 2-D. the indices of the filled sparse tensor.SparseToDense.Inputs.defaultValueScalar value to set for indices not specified insparse_indices.SparseDenseCwiseAdd.Inputs.denseR-D.SparseDenseCwiseDiv.Inputs.denseR-D.SparseDenseCwiseMul.Inputs.denseR-D.SparseSegmentMeanGrad.Inputs.denseOutputDim0dimension 0 of "data" passed to SparseSegmentMean op.SparseSegmentSqrtNGrad.Inputs.denseOutputDim0dimension 0 of "data" passed to SparseSegmentSqrtN op.SparseSegmentSumGrad.Inputs.denseOutputDim0dimension 0 of "data" passed to SparseSegmentSum op.SparseBincount.Inputs.denseShape1D int64Tensor.SparseCountSparseOutput.Inputs.denseShapeTensor containing the dense shape of the sparse tensor to count.SparseFillEmptyRows.Inputs.denseShape1-D. the shape of the sparse tensor.SparseSegmentMeanGrad.Inputs.gradgradient propagated to the SparseSegmentMean op.SparseSegmentSqrtNGrad.Inputs.gradgradient propagated to the SparseSegmentSqrtN op.SparseSegmentSumGrad.Inputs.gradgradient propagated to the SparseSegmentSum op.SparseAccumulatorApplyGradient.Inputs.gradientIndicesIndices of the sparse gradient to be accumulated.SparseAccumulatorApplyGradient.Inputs.gradientShapeShape of the sparse gradient to be accumulated.SparseAccumulatorApplyGradient.Inputs.gradientValuesValues are the non-zero slices of the gradient, and must have the same first dimension as indices, i.e., the nnz represented by indices and values must be consistent.SparseFillEmptyRowsGrad.Inputs.gradValues1-D.SparseAccumulatorApplyGradient.Inputs.handleThe handle to a accumulator.SparseAccumulatorTakeGradient.Inputs.handleThe handle to a SparseConditionalAccumulator.SparseBincount.Inputs.indices2D int64Tensor.SparseCountSparseOutput.Inputs.indicesTensor containing the indices of the sparse tensor to count.SparseFillEmptyRows.Inputs.indices2-D. the indices of the sparse tensor.SparseSegmentMean.Inputs.indicesA 1-D tensor.SparseSegmentMeanGrad.Inputs.indicesindices passed to the corresponding SparseSegmentMean op.SparseSegmentMeanWithNumSegments.Inputs.indicesA 1-D tensor.SparseSegmentSqrtN.Inputs.indicesA 1-D tensor.SparseSegmentSqrtNGrad.Inputs.indicesindices passed to the corresponding SparseSegmentSqrtN op.SparseSegmentSqrtNWithNumSegments.Inputs.indicesA 1-D tensor.SparseSegmentSum.Inputs.indicesA 1-D tensor.SparseSegmentSumGrad.Inputs.indicesindices passed to the corresponding SparseSegmentSum op.SparseSegmentSumWithNumSegments.Inputs.indicesA 1-D tensor.SparseSlice.Inputs.indices2-D tensor represents the indices of the sparse tensor.SparseSplit.Inputs.indices2-D tensor represents the indices of the sparse tensor.ConvertToListOfSparseCoreCooTensors.Inputs.indicesOrRowSplitsThe indicesOrRowSplits inputSparseReduceMax.Inputs.inputIndices2-D.SparseReduceMaxSparse.Inputs.inputIndices2-D.SparseReduceSum.Inputs.inputIndices2-D.SparseReduceSumSparse.Inputs.inputIndices2-D.SparseReorder.Inputs.inputIndices2-D.SparseReshape.Inputs.inputIndices2-D.SparseSliceGrad.Inputs.inputIndices2-D.SparseReduceMax.Inputs.inputShape1-D.SparseReduceMaxSparse.Inputs.inputShape1-D.SparseReduceSum.Inputs.inputShape1-D.SparseReduceSumSparse.Inputs.inputShape1-D.SparseReorder.Inputs.inputShape1-D.SparseReshape.Inputs.inputShape1-D.SparseSliceGrad.Inputs.inputStart1-D. tensor represents the start of the slice.SparseReduceMax.Inputs.inputValues1-D.SparseReduceMaxSparse.Inputs.inputValues1-D.SparseReduceSum.Inputs.inputValues1-D.SparseReduceSumSparse.Inputs.inputValues1-D.SparseReorder.Inputs.inputValues1-D.SparseAccumulatorApplyGradient.Inputs.localStepThe local_step value at which the sparse gradient was computed.SparseReshape.Inputs.newShape1-D.SparseCrossHashed.Inputs.numBucketsIt is used if hashed_output is true. output = hashed_value%num_buckets if num_buckets > 0 else hashed_value.SparseAccumulatorTakeGradient.Inputs.numRequiredNumber of gradients required before we return an aggregate.SparseSegmentMeanWithNumSegments.Inputs.numSegmentsShould equal the number of distinct segment IDs.SparseSegmentSqrtNWithNumSegments.Inputs.numSegmentsShould equal the number of distinct segment IDs.SparseSegmentSumWithNumSegments.Inputs.numSegmentsShould equal the number of distinct segment IDs.SparseSliceGrad.Inputs.outputIndices2-D.SparseToDense.Inputs.outputShape1-D.SparseReduceMax.Inputs.reductionAxes1-D.SparseReduceMaxSparse.Inputs.reductionAxes1-D.SparseReduceSum.Inputs.reductionAxes1-D.SparseReduceSumSparse.Inputs.reductionAxes1-D.SparseFillEmptyRowsGrad.Inputs.reverseIndexMap1-D.SparseCrossHashed.Inputs.saltSpecify the salt that will be used by the siphash function.SparseSegmentMean.Inputs.segmentIdsA 1-D tensor.SparseSegmentMeanGrad.Inputs.segmentIdssegment_ids passed to the corresponding SparseSegmentMean op.SparseSegmentMeanWithNumSegments.Inputs.segmentIdsA 1-D tensor.SparseSegmentSqrtN.Inputs.segmentIdsA 1-D tensor.SparseSegmentSqrtNGrad.Inputs.segmentIdssegment_ids passed to the corresponding SparseSegmentSqrtN op.SparseSegmentSqrtNWithNumSegments.Inputs.segmentIdsA 1-D tensor.SparseSegmentSum.Inputs.segmentIdsA 1-D tensor.SparseSegmentSumGrad.Inputs.segmentIdssegment_ids passed to the corresponding SparseSegmentSum op.SparseSegmentSumWithNumSegments.Inputs.segmentIdsA 1-D tensor.SparseCross.Inputs.sepstring used when joining a list of string inputs, can be used as separator later.DeserializeSparse.Inputs.serializedSparseThe serializedSparseTensorobjects.DenseToDenseSetOperation.Inputs.set1Tensorwith rankn. 1stn-1dimensions must be the same asset2.DenseToSparseSetOperation.Inputs.set1Tensorwith rankn. 1stn-1dimensions must be the same asset2.SparseToSparseSetOperation.Inputs.set1Indices2DTensor, indices of aSparseTensor.SparseToSparseSetOperation.Inputs.set1Shape1DTensor, shape of aSparseTensor.SparseToSparseSetOperation.Inputs.set1Values1DTensor, values of aSparseTensor.DenseToDenseSetOperation.Inputs.set2Tensorwith rankn. 1stn-1dimensions must be the same asset1.DenseToSparseSetOperation.Inputs.set2Indices2DTensor, indices of aSparseTensor.SparseToSparseSetOperation.Inputs.set2Indices2DTensor, indices of aSparseTensor.DenseToSparseSetOperation.Inputs.set2Shape1DTensor, shape of aSparseTensor.SparseToSparseSetOperation.Inputs.set2Shape1DTensor, shape of aSparseTensor.DenseToSparseSetOperation.Inputs.set2Values1DTensor, values of aSparseTensor.SparseToSparseSetOperation.Inputs.set2Values1DTensor, values of aSparseTensor.SparseSlice.Inputs.shape1-D. tensor represents the shape of the sparse tensor.SparseSplit.Inputs.shape1-D. tensor represents the shape of the sparse tensor. output indices: A list of 1-D tensors represents the indices of the output sparse tensors.SparseBincount.Inputs.sizeOutputnon-negative int scalarTensor.SparseSlice.Inputs.sizeOutput1-D. tensor represents the size of the slice. output indices: A list of 1-D tensors represents the indices of the output sparse tensors.TakeManySparseFromTensorsMap.Inputs.sparseHandles1-D, TheNserializedSparseTensorobjects.AddManySparseToTensorsMap.Inputs.sparseIndices2-D.AddSparseToTensorsMap.Inputs.sparseIndices2-D.SparseToDense.Inputs.sparseIndices0-D, 1-D, or 2-D.AddManySparseToTensorsMap.Inputs.sparseShape1-D.AddSparseToTensorsMap.Inputs.sparseShape1-D.AddManySparseToTensorsMap.Inputs.sparseValues1-D.AddSparseToTensorsMap.Inputs.sparseValues1-D.SparseToDense.Inputs.sparseValues1-D.SparseDenseCwiseAdd.Inputs.spIndices2-D.SparseDenseCwiseDiv.Inputs.spIndices2-D.SparseDenseCwiseMul.Inputs.spIndices2-D.SparseSoftmax.Inputs.spIndices2-D.SparseSplit.Inputs.splitDim0-D.ConvertToSparseCoreCsrWrappedCooTensor.Inputs.splitsThe splits inputSparseDenseCwiseAdd.Inputs.spShape1-D.SparseDenseCwiseDiv.Inputs.spShape1-D.SparseDenseCwiseMul.Inputs.spShape1-D.SparseSoftmax.Inputs.spShape1-D.SparseDenseCwiseAdd.Inputs.spValues1-D.SparseDenseCwiseDiv.Inputs.spValues1-D.SparseDenseCwiseMul.Inputs.spValues1-D.SparseSoftmax.Inputs.spValues1-D.SparseSlice.Inputs.start1-D. tensor represents the start of the slice.SparseCrossHashed.Inputs.strongHashboolean, if true, siphash with salt will be used instead of farmhash.SparseAddGrad.Inputs.sumIndices2-D.SparseAdd.Inputs.thresh0-D.ConvertToListOfSparseCoreCooTensors.Inputs.valuesThe values inputDenseCountSparseOutput.Inputs.valuesTensor containing data to count.SparseBincount.Inputs.values1D intTensor.SparseCountSparseOutput.Inputs.valuesTensor containing values of the sparse tensor to count.SparseFillEmptyRows.Inputs.values1-D. the values of the sparse tensor.SparseSlice.Inputs.values1-D tensor represents the values of the sparse tensor.SparseSplit.Inputs.values1-D tensor represents the values of the sparse tensor.ConvertToListOfSparseCoreCooTensors.Inputs.weightsThe weights inputDenseCountSparseOutput.Inputs.weightsA Tensor of the same shape as indices containing per-index weight values.SparseBincount.Inputs.weightsis an int32, int64, float32, or float64Tensorwith the same shape asinput, or a length-0Tensor, in which case it acts as all weights equal to 1.SparseCountSparseOutput.Inputs.weightsA Tensor of the same shape as indices containing per-index weight values.Fields in org.tensorflow.op.sparse with type parameters of type OperandModifier and TypeFieldDescriptionGetStatsFromListOfSparseCoreCooTensors.Inputs.colIdsListThe colIdsList inputSortListOfSparseCoreCooTensors.Inputs.colIdsListThe colIdsList inputSparseCross.Inputs.denseInputs2-D.SparseCrossHashed.Inputs.denseInputs2-D.GetStatsFromListOfSparseCoreCooTensors.Inputs.gainsListThe gainsList inputSortListOfSparseCoreCooTensors.Inputs.gainsListThe gainsList inputConvertToSparseCoreCsrWrappedCooTensor.Inputs.idCountsListThe idCountsList inputSparseConcat.Inputs.indices2-D.SparseCross.Inputs.indices2-D.SparseCrossHashed.Inputs.indices2-D.GetStatsFromListOfSparseCoreCooTensors.Inputs.rowIdsListThe rowIdsList inputSortListOfSparseCoreCooTensors.Inputs.rowIdsListThe rowIdsList inputSparseConcat.Inputs.shapes1-D.SparseCross.Inputs.shapes1-D.SparseCrossHashed.Inputs.shapes1-D.ConvertToSparseCoreCsrWrappedCooTensor.Inputs.sortedColIdsListThe sortedColIdsList inputConvertToSparseCoreCsrWrappedCooTensor.Inputs.sortedGainsListThe sortedGainsList inputConvertToSparseCoreCsrWrappedCooTensor.Inputs.sortedRowIdsListThe sortedRowIdsList inputSparseConcat.Inputs.values1-D.SparseCross.Inputs.values1-D.SparseCrossHashed.Inputs.values1-D.Methods in org.tensorflow.op.sparse with parameters of type OperandModifier and TypeMethodDescriptionstatic AddManySparseToTensorsMapAddManySparseToTensorsMap.create(Scope scope, Operand<TInt64> sparseIndices, Operand<? extends TType> sparseValues, Operand<TInt64> sparseShape, AddManySparseToTensorsMap.Options... options) Factory method to create a class wrapping a new AddManySparseToTensorsMap operation.static AddSparseToTensorsMapAddSparseToTensorsMap.create(Scope scope, Operand<TInt64> sparseIndices, Operand<? extends TType> sparseValues, Operand<TInt64> sparseShape, AddSparseToTensorsMap.Options... options) Factory method to create a class wrapping a new AddSparseToTensorsMap operation.ConvertToListOfSparseCoreCooTensors.create(Scope scope, Operand<TInt32> indicesOrRowSplits, Operand<TInt32> values, Operand<TFloat32> weights, Long sampleCount, Long numScPerChip, Long rowOffset, Long colOffset, Long colShift, Long numScShards, Long stackedTableSampleCount, String combiner) Factory method to create a class wrapping a new ConvertToListOfSparseCoreCooTensors operation.ConvertToSparseCoreCsrWrappedCooTensor.create(Scope scope, Iterable<Operand<TInt32>> sortedRowIdsList, Iterable<Operand<TInt32>> sortedColIdsList, Iterable<Operand<TFloat32>> sortedGainsList, Iterable<Operand<TInt32>> idCountsList, Operand<TInt64> splits, Long sampleCountPerSc, Long numReplica, Long maxMinibatchesPerSc, Long maxIdsPerChipPerSample, Long tableVocabSize, Long featureWidth, String tableName, Boolean allowIdDropping) Factory method to create a class wrapping a new ConvertToSparseCoreCsrWrappedCooTensor operation.static <U extends TNumber>
DenseCountSparseOutput<U> DenseCountSparseOutput.create(Scope scope, Operand<? extends TNumber> values, Operand<U> weights, Boolean binaryOutput, DenseCountSparseOutput.Options... options) Factory method to create a class wrapping a new DenseCountSparseOutput operation.static <T extends TType>
DenseToDenseSetOperation<T> DenseToDenseSetOperation.create(Scope scope, Operand<T> set1, Operand<T> set2, String setOperation, DenseToDenseSetOperation.Options... options) Factory method to create a class wrapping a new DenseToDenseSetOperation operation.static <T extends TType>
DenseToSparseSetOperation<T> DenseToSparseSetOperation.create(Scope scope, Operand<T> set1, Operand<TInt64> set2Indices, Operand<T> set2Values, Operand<TInt64> set2Shape, String setOperation, DenseToSparseSetOperation.Options... options) Factory method to create a class wrapping a new DenseToSparseSetOperation operation.static <U extends TType>
DeserializeSparse<U> Factory method to create a class wrapping a new DeserializeSparse operation.SparseAccumulatorApplyGradient.create(Scope scope, Operand<TString> handle, Operand<TInt64> localStep, Operand<TInt64> gradientIndices, Operand<? extends TType> gradientValues, Operand<TInt64> gradientShape, Boolean hasKnownShape) Factory method to create a class wrapping a new SparseAccumulatorApplyGradient operation.static <T extends TType>
SparseAccumulatorTakeGradient<T> SparseAccumulatorTakeGradient.create(Scope scope, Operand<TString> handle, Operand<TInt32> numRequired, Class<T> dtype) Factory method to create a class wrapping a new SparseAccumulatorTakeGradient operation.SparseAdd.create(Scope scope, Operand<TInt64> aIndices, Operand<T> aValues, Operand<TInt64> aShape, Operand<TInt64> bIndices, Operand<T> bValues, Operand<TInt64> bShape, Operand<? extends TNumber> thresh) Factory method to create a class wrapping a new SparseAdd operation.static <T extends TType>
SparseAddGrad<T> SparseAddGrad.create(Scope scope, Operand<T> backpropValGrad, Operand<TInt64> aIndices, Operand<TInt64> bIndices, Operand<TInt64> sumIndices) Factory method to create a class wrapping a new SparseAddGrad operation.static <U extends TNumber, T extends TNumber>
SparseBincount<U> SparseBincount.create(Scope scope, Operand<TInt64> indices, Operand<T> values, Operand<TInt64> denseShape, Operand<T> sizeOutput, Operand<U> weights, SparseBincount.Options... options) Factory method to create a class wrapping a new SparseBincount operation.static <U extends TNumber>
SparseCountSparseOutput<U> SparseCountSparseOutput.create(Scope scope, Operand<TInt64> indices, Operand<? extends TNumber> values, Operand<TInt64> denseShape, Operand<U> weights, Boolean binaryOutput, SparseCountSparseOutput.Options... options) Factory method to create a class wrapping a new SparseCountSparseOutput operation.static SparseCrossSparseCross.create(Scope scope, Iterable<Operand<TInt64>> indices, Iterable<Operand<?>> values, Iterable<Operand<TInt64>> shapes, Iterable<Operand<?>> denseInputs, Operand<TString> sep) Factory method to create a class wrapping a new SparseCrossV2 operation.static SparseCrossHashedSparseCrossHashed.create(Scope scope, Iterable<Operand<TInt64>> indices, Iterable<Operand<?>> values, Iterable<Operand<TInt64>> shapes, Iterable<Operand<?>> denseInputs, Operand<TInt64> numBuckets, Operand<TBool> strongHash, Operand<TInt64> salt) Factory method to create a class wrapping a new SparseCrossHashed operation.static <T extends TType>
SparseDenseCwiseAdd<T> SparseDenseCwiseAdd.create(Scope scope, Operand<TInt64> spIndices, Operand<T> spValues, Operand<TInt64> spShape, Operand<T> dense) Factory method to create a class wrapping a new SparseDenseCwiseAdd operation.static <T extends TType>
SparseDenseCwiseDiv<T> SparseDenseCwiseDiv.create(Scope scope, Operand<TInt64> spIndices, Operand<T> spValues, Operand<TInt64> spShape, Operand<T> dense) Factory method to create a class wrapping a new SparseDenseCwiseDiv operation.static <T extends TType>
SparseDenseCwiseMul<T> SparseDenseCwiseMul.create(Scope scope, Operand<TInt64> spIndices, Operand<T> spValues, Operand<TInt64> spShape, Operand<T> dense) Factory method to create a class wrapping a new SparseDenseCwiseMul operation.static <T extends TType>
SparseFillEmptyRows<T> SparseFillEmptyRows.create(Scope scope, Operand<TInt64> indices, Operand<T> values, Operand<TInt64> denseShape, Operand<T> defaultValue) Factory method to create a class wrapping a new SparseFillEmptyRows operation.static <T extends TType>
SparseFillEmptyRowsGrad<T> Factory method to create a class wrapping a new SparseFillEmptyRowsGrad operation.static SparseMatMulSparseMatMul.create(Scope scope, Operand<? extends TNumber> a, Operand<? extends TNumber> b, SparseMatMul.Options... options) Factory method to create a class wrapping a new SparseMatMul operation.static <T extends TNumber>
SparseReduceMax<T> SparseReduceMax.create(Scope scope, Operand<TInt64> inputIndices, Operand<T> inputValues, Operand<TInt64> inputShape, Operand<TInt32> reductionAxes, SparseReduceMax.Options... options) Factory method to create a class wrapping a new SparseReduceMax operation.static <T extends TNumber>
SparseReduceMaxSparse<T> SparseReduceMaxSparse.create(Scope scope, Operand<TInt64> inputIndices, Operand<T> inputValues, Operand<TInt64> inputShape, Operand<TInt32> reductionAxes, SparseReduceMaxSparse.Options... options) Factory method to create a class wrapping a new SparseReduceMaxSparse operation.static <T extends TType>
SparseReduceSum<T> SparseReduceSum.create(Scope scope, Operand<TInt64> inputIndices, Operand<T> inputValues, Operand<TInt64> inputShape, Operand<TInt32> reductionAxes, SparseReduceSum.Options... options) Factory method to create a class wrapping a new SparseReduceSum operation.static <T extends TType>
SparseReduceSumSparse<T> SparseReduceSumSparse.create(Scope scope, Operand<TInt64> inputIndices, Operand<T> inputValues, Operand<TInt64> inputShape, Operand<TInt32> reductionAxes, SparseReduceSumSparse.Options... options) Factory method to create a class wrapping a new SparseReduceSumSparse operation.static <T extends TType>
SparseReorder<T> SparseReorder.create(Scope scope, Operand<TInt64> inputIndices, Operand<T> inputValues, Operand<TInt64> inputShape) Factory method to create a class wrapping a new SparseReorder operation.static SparseReshapeSparseReshape.create(Scope scope, Operand<TInt64> inputIndices, Operand<TInt64> inputShape, Operand<TInt64> newShape) Factory method to create a class wrapping a new SparseReshape operation.static <T extends TNumber>
SparseSegmentMean<T> SparseSegmentMean.create(Scope scope, Operand<T> data, Operand<? extends TNumber> indices, Operand<? extends TNumber> segmentIds, SparseSegmentMean.Options... options) Factory method to create a class wrapping a new SparseSegmentMean operation.static <T extends TNumber, U extends TNumber>
SparseSegmentMeanGrad<T, U> SparseSegmentMeanGrad.create(Scope scope, Operand<T> grad, Operand<U> indices, Operand<? extends TNumber> segmentIds, Operand<TInt32> denseOutputDim0) Factory method to create a class wrapping a new SparseSegmentMeanGradV2 operation.static <T extends TNumber>
SparseSegmentMeanWithNumSegments<T> SparseSegmentMeanWithNumSegments.create(Scope scope, Operand<T> data, Operand<? extends TNumber> indices, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments, SparseSegmentMeanWithNumSegments.Options... options) Factory method to create a class wrapping a new SparseSegmentMeanWithNumSegments operation.static <T extends TNumber>
SparseSegmentSqrtN<T> SparseSegmentSqrtN.create(Scope scope, Operand<T> data, Operand<? extends TNumber> indices, Operand<? extends TNumber> segmentIds, SparseSegmentSqrtN.Options... options) Factory method to create a class wrapping a new SparseSegmentSqrtN operation.static <T extends TNumber, U extends TNumber>
SparseSegmentSqrtNGrad<T, U> SparseSegmentSqrtNGrad.create(Scope scope, Operand<T> grad, Operand<U> indices, Operand<? extends TNumber> segmentIds, Operand<TInt32> denseOutputDim0) Factory method to create a class wrapping a new SparseSegmentSqrtNGradV2 operation.static <T extends TNumber>
SparseSegmentSqrtNWithNumSegments<T> SparseSegmentSqrtNWithNumSegments.create(Scope scope, Operand<T> data, Operand<? extends TNumber> indices, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments, SparseSegmentSqrtNWithNumSegments.Options... options) Factory method to create a class wrapping a new SparseSegmentSqrtNWithNumSegments operation.static <T extends TNumber>
SparseSegmentSum<T> SparseSegmentSum.create(Scope scope, Operand<T> data, Operand<? extends TNumber> indices, Operand<? extends TNumber> segmentIds, SparseSegmentSum.Options... options) Factory method to create a class wrapping a new SparseSegmentSum operation.static <T extends TNumber, U extends TNumber>
SparseSegmentSumGrad<T, U> SparseSegmentSumGrad.create(Scope scope, Operand<T> grad, Operand<U> indices, Operand<? extends TNumber> segmentIds, Operand<TInt32> denseOutputDim0) Factory method to create a class wrapping a new SparseSegmentSumGradV2 operation.static <T extends TNumber>
SparseSegmentSumWithNumSegments<T> SparseSegmentSumWithNumSegments.create(Scope scope, Operand<T> data, Operand<? extends TNumber> indices, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments, SparseSegmentSumWithNumSegments.Options... options) Factory method to create a class wrapping a new SparseSegmentSumWithNumSegments operation.static <T extends TType>
SparseSlice<T> SparseSlice.create(Scope scope, Operand<TInt64> indices, Operand<T> values, Operand<TInt64> shape, Operand<TInt64> start, Operand<TInt64> sizeOutput) Factory method to create a class wrapping a new SparseSlice operation.static <T extends TType>
SparseSliceGrad<T> SparseSliceGrad.create(Scope scope, Operand<T> backpropValGrad, Operand<TInt64> inputIndices, Operand<TInt64> inputStart, Operand<TInt64> outputIndices) Factory method to create a class wrapping a new SparseSliceGrad operation.static <T extends TNumber>
SparseSoftmax<T> SparseSoftmax.create(Scope scope, Operand<TInt64> spIndices, Operand<T> spValues, Operand<TInt64> spShape) Factory method to create a class wrapping a new SparseSoftmax operation.static <T extends TNumber>
SparseSparseMaximum<T> SparseSparseMaximum.create(Scope scope, Operand<TInt64> aIndices, Operand<T> aValues, Operand<TInt64> aShape, Operand<TInt64> bIndices, Operand<T> bValues, Operand<TInt64> bShape) Factory method to create a class wrapping a new SparseSparseMaximum operation.static <T extends TType>
SparseSparseMinimum<T> SparseSparseMinimum.create(Scope scope, Operand<TInt64> aIndices, Operand<T> aValues, Operand<TInt64> aShape, Operand<TInt64> bIndices, Operand<T> bValues, Operand<TInt64> bShape) Factory method to create a class wrapping a new SparseSparseMinimum operation.static <T extends TType>
SparseSplit<T> SparseSplit.create(Scope scope, Operand<TInt64> splitDim, Operand<TInt64> indices, Operand<T> values, Operand<TInt64> shape, Long numSplit) Factory method to create a class wrapping a new SparseSplit operation.static <U extends TType, T extends TNumber>
SparseTensorDenseAdd<U> SparseTensorDenseAdd.create(Scope scope, Operand<T> aIndices, Operand<U> aValues, Operand<T> aShape, Operand<U> b) Factory method to create a class wrapping a new SparseTensorDenseAdd operation.static <U extends TType>
SparseTensorDenseMatMul<U> SparseTensorDenseMatMul.create(Scope scope, Operand<? extends TNumber> aIndices, Operand<U> aValues, Operand<TInt64> aShape, Operand<U> b, SparseTensorDenseMatMul.Options... options) Factory method to create a class wrapping a new SparseTensorDenseMatMul operation.static <U extends TType, T extends TNumber>
SparseToDense<U> SparseToDense.create(Scope scope, Operand<T> sparseIndices, Operand<T> outputShape, Operand<U> sparseValues, Operand<U> defaultValue, SparseToDense.Options... options) Factory method to create a class wrapping a new SparseToDense operation.static <T extends TType>
SparseToSparseSetOperation<T> SparseToSparseSetOperation.create(Scope scope, Operand<TInt64> set1Indices, Operand<T> set1Values, Operand<TInt64> set1Shape, Operand<TInt64> set2Indices, Operand<T> set2Values, Operand<TInt64> set2Shape, String setOperation, SparseToSparseSetOperation.Options... options) Factory method to create a class wrapping a new SparseToSparseSetOperation operation.static <T extends TType>
TakeManySparseFromTensorsMap<T> TakeManySparseFromTensorsMap.create(Scope scope, Operand<TInt64> sparseHandles, Class<T> dtype, TakeManySparseFromTensorsMap.Options... options) Factory method to create a class wrapping a new TakeManySparseFromTensorsMap operation.Method parameters in org.tensorflow.op.sparse with type arguments of type OperandModifier and TypeMethodDescriptionConvertToSparseCoreCsrWrappedCooTensor.create(Scope scope, Iterable<Operand<TInt32>> sortedRowIdsList, Iterable<Operand<TInt32>> sortedColIdsList, Iterable<Operand<TFloat32>> sortedGainsList, Iterable<Operand<TInt32>> idCountsList, Operand<TInt64> splits, Long sampleCountPerSc, Long numReplica, Long maxMinibatchesPerSc, Long maxIdsPerChipPerSample, Long tableVocabSize, Long featureWidth, String tableName, Boolean allowIdDropping) Factory method to create a class wrapping a new ConvertToSparseCoreCsrWrappedCooTensor operation.GetStatsFromListOfSparseCoreCooTensors.create(Scope scope, Iterable<Operand<TInt32>> rowIdsList, Iterable<Operand<TInt32>> colIdsList, Iterable<Operand<TFloat32>> gainsList, List<Long> sampleCountList, List<Long> colOffsetList, Long numReplica, Long tableVocabSize, Long featureWidth, Long numScPerChip, String tableName) Factory method to create a class wrapping a new GetStatsFromListOfSparseCoreCooTensors operation.SortListOfSparseCoreCooTensors.create(Scope scope, Iterable<Operand<TInt32>> rowIdsList, Iterable<Operand<TInt32>> colIdsList, Iterable<Operand<TFloat32>> gainsList, List<Long> sampleCountList, List<Long> colOffsetList, Long numReplica, Long tableVocabSize, Long featureWidth, Long numScPerChip, Long maxIdsPerSparseCore, Long maxUniqueIdsPerSparseCore, String tableName) Factory method to create a class wrapping a new SortListOfSparseCoreCooTensors operation.static <T extends TType>
SparseConcat<T> SparseConcat.create(Scope scope, Iterable<Operand<TInt64>> indices, Iterable<Operand<T>> values, Iterable<Operand<TInt64>> shapes, Long concatDim) Factory method to create a class wrapping a new SparseConcat operation.static SparseCrossSparseCross.create(Scope scope, Iterable<Operand<TInt64>> indices, Iterable<Operand<?>> values, Iterable<Operand<TInt64>> shapes, Iterable<Operand<?>> denseInputs, Operand<TString> sep) Factory method to create a class wrapping a new SparseCrossV2 operation.static SparseCrossHashedSparseCrossHashed.create(Scope scope, Iterable<Operand<TInt64>> indices, Iterable<Operand<?>> values, Iterable<Operand<TInt64>> shapes, Iterable<Operand<?>> denseInputs, Operand<TInt64> numBuckets, Operand<TBool> strongHash, Operand<TInt64> salt) Factory method to create a class wrapping a new SparseCrossHashed operation. -
Uses of Operand in org.tensorflow.op.strings
Classes in org.tensorflow.op.strings that implement OperandModifier and TypeClassDescriptionfinal classJoins the strings in the given list of string tensors into one tensor; with the given separator (default is an empty separator).final classConverts all uppercase characters into their respective lowercase replacements.final classJoins a string Tensor across the given dimensions.final classCheck if the input matches the regex pattern.final classReplaces matches of thepatternregular expression ininputwith the replacement string provided inrewrite.final classCheck if the input matches the regex pattern.final classReplaces the match of pattern in input with rewrite.final classFormats a string template using a list of tensors.final classString lengths ofinput.final classStrip leading and trailing whitespaces from the Tensor.final classReturn substrings fromTensorof strings.final classConverts each string in the input Tensor to its hash mod by a number of buckets.final classConverts each string in the input Tensor to its hash mod by a number of buckets.final classConverts each string in the input Tensor to its hash mod by a number of buckets.final classConverts each string in the input Tensor to the specified numeric type.final classEncode a tensor of ints into unicode strings.final classDetermine the script codes of a given tensor of Unicode integer code points.final classTranscode the input text from a source encoding to a destination encoding.final classThe UnsortedSegmentJoin operationfinal classConverts all lowercase characters into their respective uppercase replacements.Fields in org.tensorflow.op.strings declared as OperandModifier and TypeFieldDescriptionStringNGrams.Inputs.dataThe values tensor of the ragged string tensor to make ngrams out of.StringNGrams.Inputs.dataSplitsThe splits tensor of the ragged string tensor to make ngrams out of.Lower.Inputs.inputThe input to be lower-cased.RegexFullMatch.Inputs.inputA string tensor of the text to be processed.RegexReplace.Inputs.inputThe text to be processed.StaticRegexFullMatch.Inputs.inputA string tensor of the text to be processed.StaticRegexReplace.Inputs.inputThe text to be processed.StringLength.Inputs.inputThe strings for which to compute the length for each element.StringSplit.Inputs.input1-DstringTensor, the strings to split.Strip.Inputs.inputA stringTensorof any shape.Substr.Inputs.inputTensor of stringsToHashBucketFast.Inputs.inputThe strings to assign a hash bucket.ToHashBucketStrong.Inputs.inputThe strings to assign a hash bucket.UnicodeDecode.Inputs.inputThe text to be decoded.UnicodeDecodeWithOffsets.Inputs.inputThe text to be decoded.UnicodeScript.Inputs.inputA Tensor of int32 Unicode code points.UnicodeTranscode.Inputs.inputThe text to be processed.Upper.Inputs.inputThe input to be upper-cased.ReduceJoin.Inputs.inputsThe input to be joined.UnsortedSegmentJoin.Inputs.inputsThe inputs inputUnicodeEncode.Inputs.inputSplitsA 1D tensor specifying how the unicode codepoints should be split into strings.UnicodeEncode.Inputs.inputValuesA 1D tensor containing the unicode codepoints that should be encoded.Substr.Inputs.lenScalar defining the number of characters to include in each substringUnsortedSegmentJoin.Inputs.numSegmentsThe numSegments inputRegexFullMatch.Inputs.patternA scalar string tensor containing the regular expression to match the input.RegexReplace.Inputs.patternThe regular expression to be matched in theinputstrings.Substr.Inputs.posScalar defining the position of first character in each substringReduceJoin.Inputs.reductionIndicesThe dimensions to reduce over.RegexReplace.Inputs.rewriteThe rewrite string to be substituted for thepatternexpression where it is matched in theinputstrings.UnsortedSegmentJoin.Inputs.segmentIdsThe segmentIds inputStringSplit.Inputs.sep0-DstringTensor, the delimiter character.ToHashBucket.Inputs.stringTensorThe stringTensor inputToNumber.Inputs.stringTensorThe stringTensor inputFields in org.tensorflow.op.strings with type parameters of type OperandModifier and TypeFieldDescriptionJoin.Inputs.inputsA list of string tensors.StringFormat.Inputs.inputsThe list of tensors to format into the placeholder string.Methods in org.tensorflow.op.strings with parameters of type OperandModifier and TypeMethodDescriptionstatic LowerLower.create(Scope scope, Operand<TString> input, Lower.Options... options) Factory method to create a class wrapping a new StringLower operation.static ReduceJoinReduceJoin.create(Scope scope, Operand<TString> inputs, Operand<TInt32> reductionIndices, ReduceJoin.Options... options) Factory method to create a class wrapping a new ReduceJoin operation.static RegexFullMatchFactory method to create a class wrapping a new RegexFullMatch operation.static RegexReplaceRegexReplace.create(Scope scope, Operand<TString> input, Operand<TString> pattern, Operand<TString> rewrite, RegexReplace.Options... options) Factory method to create a class wrapping a new RegexReplace operation.static StaticRegexFullMatchFactory method to create a class wrapping a new StaticRegexFullMatch operation.static StaticRegexReplaceStaticRegexReplace.create(Scope scope, Operand<TString> input, String pattern, String rewrite, StaticRegexReplace.Options... options) Factory method to create a class wrapping a new StaticRegexReplace operation.static StringLengthStringLength.create(Scope scope, Operand<TString> input, StringLength.Options... options) Factory method to create a class wrapping a new StringLength operation.static <T extends TNumber>
StringNGrams<T> StringNGrams.create(Scope scope, Operand<TString> data, Operand<T> dataSplits, String separator, List<Long> ngramWidths, String leftPad, String rightPad, Long padWidth, Boolean preserveShortSequences) Factory method to create a class wrapping a new StringNGrams operation.static StringSplitStringSplit.create(Scope scope, Operand<TString> input, Operand<TString> sep, StringSplit.Options... options) Factory method to create a class wrapping a new StringSplitV2 operation.static StripFactory method to create a class wrapping a new StringStrip operation.Substr.create(Scope scope, Operand<TString> input, Operand<T> pos, Operand<T> len, Substr.Options... options) Factory method to create a class wrapping a new Substr operation.static ToHashBucketFactory method to create a class wrapping a new StringToHashBucket operation.static ToHashBucketFastFactory method to create a class wrapping a new StringToHashBucketFast operation.static ToHashBucketStrongFactory method to create a class wrapping a new StringToHashBucketStrong operation.Factory method to create a class wrapping a new StringToNumber operation, with the default output types.Factory method to create a class wrapping a new StringToNumber operation.static <T extends TNumber>
UnicodeDecode<T> UnicodeDecode.create(Scope scope, Operand<TString> input, String inputEncoding, Class<T> Tsplits, UnicodeDecode.Options... options) Factory method to create a class wrapping a new UnicodeDecode operation.static UnicodeDecode<TInt64> UnicodeDecode.create(Scope scope, Operand<TString> input, String inputEncoding, UnicodeDecode.Options[] options) Factory method to create a class wrapping a new UnicodeDecode operation, with the default output types.static <T extends TNumber>
UnicodeDecodeWithOffsets<T> UnicodeDecodeWithOffsets.create(Scope scope, Operand<TString> input, String inputEncoding, Class<T> Tsplits, UnicodeDecodeWithOffsets.Options... options) Factory method to create a class wrapping a new UnicodeDecodeWithOffsets operation.static UnicodeDecodeWithOffsets<TInt64> UnicodeDecodeWithOffsets.create(Scope scope, Operand<TString> input, String inputEncoding, UnicodeDecodeWithOffsets.Options[] options) Factory method to create a class wrapping a new UnicodeDecodeWithOffsets operation, with the default output types.static UnicodeEncodeUnicodeEncode.create(Scope scope, Operand<TInt32> inputValues, Operand<? extends TNumber> inputSplits, String outputEncoding, UnicodeEncode.Options... options) Factory method to create a class wrapping a new UnicodeEncode operation.static UnicodeScriptFactory method to create a class wrapping a new UnicodeScript operation.static UnicodeTranscodeUnicodeTranscode.create(Scope scope, Operand<TString> input, String inputEncoding, String outputEncoding, UnicodeTranscode.Options... options) Factory method to create a class wrapping a new UnicodeTranscode operation.static UnsortedSegmentJoinUnsortedSegmentJoin.create(Scope scope, Operand<TString> inputs, Operand<? extends TNumber> segmentIds, Operand<? extends TNumber> numSegments, UnsortedSegmentJoin.Options... options) Factory method to create a class wrapping a new UnsortedSegmentJoin operation.static UpperUpper.create(Scope scope, Operand<TString> input, Upper.Options... options) Factory method to create a class wrapping a new StringUpper operation.Method parameters in org.tensorflow.op.strings with type arguments of type OperandModifier and TypeMethodDescriptionstatic JoinFactory method to create a class wrapping a new StringJoin operation.static StringFormatStringFormat.create(Scope scope, Iterable<Operand<?>> inputs, StringFormat.Options... options) Factory method to create a class wrapping a new StringFormat operation. -
Uses of Operand in org.tensorflow.op.summary
Classes in org.tensorflow.op.summary that implement OperandModifier and TypeClassDescriptionfinal classOutputs aSummaryprotocol buffer with audio.final classOutputs aSummaryprotocol buffer with a histogram.final classOutputs aSummaryprotocol buffer with images.final classMerges summaries.final classOutputs aSummaryprotocol buffer with scalar values.final classProduces a summary of any statistics recorded by the given statistics manager.final classThe SummaryWriter operationfinal classOutputs aSummaryprotocol buffer with a tensor and per-plugin data.Fields in org.tensorflow.op.summary declared as OperandModifier and TypeFieldDescriptionWriteImageSummary.Inputs.badColorThe badColor inputCreateSummaryDbWriter.Inputs.dbUriThe dbUri inputImportEvent.Inputs.eventThe event inputCreateSummaryDbWriter.Inputs.experimentNameThe experimentName inputCreateSummaryFileWriter.Inputs.filenameSuffixThe filenameSuffix inputCreateSummaryFileWriter.Inputs.flushMillisThe flushMillis inputStatsAggregatorSummary.Inputs.iteratorThe iterator inputCreateSummaryFileWriter.Inputs.logdirThe logdir inputCreateSummaryFileWriter.Inputs.maxQueueThe maxQueue inputCreateSummaryDbWriter.Inputs.runNameThe runName inputAudioSummary.Inputs.sampleRateThe sample rate of the signal in hertz.WriteAudioSummary.Inputs.sampleRateThe sampleRate inputTensorSummary.Inputs.serializedSummaryMetadataA serialized SummaryMetadata proto.WriteAudioSummary.Inputs.stepThe step inputWriteGraphSummary.Inputs.stepThe step inputWriteHistogramSummary.Inputs.stepThe step inputWriteImageSummary.Inputs.stepThe step inputWriteRawProtoSummary.Inputs.stepThe step inputWriteScalarSummary.Inputs.stepThe step inputWriteSummary.Inputs.stepThe step inputWriteSummary.Inputs.summaryMetadataThe summaryMetadata inputAudioSummary.Inputs.tagScalar.HistogramSummary.Inputs.tagScalar.ImageSummary.Inputs.tagScalar.TensorSummary.Inputs.tagA string attached to this summary.WriteAudioSummary.Inputs.tagThe tag inputWriteHistogramSummary.Inputs.tagThe tag inputWriteImageSummary.Inputs.tagThe tag inputWriteScalarSummary.Inputs.tagThe tag inputWriteSummary.Inputs.tagThe tag inputScalarSummary.Inputs.tagsTags for the summary.AudioSummary.Inputs.tensor2-D of shape[batch_size, frames].ImageSummary.Inputs.tensor4-D of shape[batch_size, height, width, channels]wherechannelsis 1, 3, or 4.TensorSummary.Inputs.tensorA tensor to serialize.WriteAudioSummary.Inputs.tensorThe tensor inputWriteGraphSummary.Inputs.tensorThe tensor inputWriteImageSummary.Inputs.tensorThe tensor inputWriteRawProtoSummary.Inputs.tensorThe tensor inputWriteSummary.Inputs.tensorThe tensor inputCreateSummaryDbWriter.Inputs.userNameThe userName inputWriteScalarSummary.Inputs.valueThe value inputHistogramSummary.Inputs.valuesAny shape.ScalarSummary.Inputs.valuesSame shape as `tags.WriteHistogramSummary.Inputs.valuesThe values inputCloseSummaryWriter.Inputs.writerThe writer inputCreateSummaryDbWriter.Inputs.writerThe writer inputCreateSummaryFileWriter.Inputs.writerThe writer inputFlushSummaryWriter.Inputs.writerThe writer inputImportEvent.Inputs.writerThe writer inputWriteAudioSummary.Inputs.writerThe writer inputWriteGraphSummary.Inputs.writerThe writer inputWriteHistogramSummary.Inputs.writerThe writer inputWriteImageSummary.Inputs.writerThe writer inputWriteRawProtoSummary.Inputs.writerThe writer inputWriteScalarSummary.Inputs.writerThe writer inputWriteSummary.Inputs.writerThe writer inputFields in org.tensorflow.op.summary with type parameters of type OperandMethods in org.tensorflow.op.summary with parameters of type OperandModifier and TypeMethodDescriptionstatic AudioSummaryAudioSummary.create(Scope scope, Operand<TString> tag, Operand<TFloat32> tensor, Operand<TFloat32> sampleRate, AudioSummary.Options... options) Factory method to create a class wrapping a new AudioSummaryV2 operation.static CloseSummaryWriterFactory method to create a class wrapping a new CloseSummaryWriter operation.static CreateSummaryDbWriterCreateSummaryDbWriter.create(Scope scope, Operand<? extends TType> writer, Operand<TString> dbUri, Operand<TString> experimentName, Operand<TString> runName, Operand<TString> userName) Factory method to create a class wrapping a new CreateSummaryDbWriter operation.static CreateSummaryFileWriterCreateSummaryFileWriter.create(Scope scope, Operand<? extends TType> writer, Operand<TString> logdir, Operand<TInt32> maxQueue, Operand<TInt32> flushMillis, Operand<TString> filenameSuffix) Factory method to create a class wrapping a new CreateSummaryFileWriter operation.static FlushSummaryWriterFactory method to create a class wrapping a new FlushSummaryWriter operation.static HistogramSummaryFactory method to create a class wrapping a new HistogramSummary operation.static ImageSummaryImageSummary.create(Scope scope, Operand<TString> tag, Operand<? extends TNumber> tensor, ImageSummary.Options... options) Factory method to create a class wrapping a new ImageSummary operation.static ImportEventFactory method to create a class wrapping a new ImportEvent operation.static ScalarSummaryFactory method to create a class wrapping a new ScalarSummary operation.static StatsAggregatorSummaryFactory method to create a class wrapping a new StatsAggregatorSummary operation.static TensorSummaryTensorSummary.create(Scope scope, Operand<TString> tag, Operand<? extends TType> tensor, Operand<TString> serializedSummaryMetadata) Factory method to create a class wrapping a new TensorSummaryV2 operation.static WriteAudioSummaryWriteAudioSummary.create(Scope scope, Operand<? extends TType> writer, Operand<TInt64> step, Operand<TString> tag, Operand<TFloat32> tensor, Operand<TFloat32> sampleRate, WriteAudioSummary.Options... options) Factory method to create a class wrapping a new WriteAudioSummary operation.static WriteGraphSummaryWriteGraphSummary.create(Scope scope, Operand<? extends TType> writer, Operand<TInt64> step, Operand<TString> tensor) Factory method to create a class wrapping a new WriteGraphSummary operation.static WriteHistogramSummaryWriteHistogramSummary.create(Scope scope, Operand<? extends TType> writer, Operand<TInt64> step, Operand<TString> tag, Operand<? extends TType> values) Factory method to create a class wrapping a new WriteHistogramSummary operation.static WriteImageSummaryWriteImageSummary.create(Scope scope, Operand<? extends TType> writer, Operand<TInt64> step, Operand<TString> tag, Operand<? extends TNumber> tensor, Operand<TUint8> badColor, WriteImageSummary.Options... options) Factory method to create a class wrapping a new WriteImageSummary operation.static WriteRawProtoSummaryWriteRawProtoSummary.create(Scope scope, Operand<? extends TType> writer, Operand<TInt64> step, Operand<TString> tensor) Factory method to create a class wrapping a new WriteRawProtoSummary operation.static WriteScalarSummaryWriteScalarSummary.create(Scope scope, Operand<? extends TType> writer, Operand<TInt64> step, Operand<TString> tag, Operand<? extends TNumber> value) Factory method to create a class wrapping a new WriteScalarSummary operation.static WriteSummaryWriteSummary.create(Scope scope, Operand<? extends TType> writer, Operand<TInt64> step, Operand<? extends TType> tensor, Operand<TString> tag, Operand<TString> summaryMetadata) Factory method to create a class wrapping a new WriteSummary operation.Method parameters in org.tensorflow.op.summary with type arguments of type Operand -
Uses of Operand in org.tensorflow.op.tpu
Classes in org.tensorflow.op.tpu that implement OperandModifier and TypeClassDescriptionfinal classAn Op to exchange data across TPU replicas.final classAn op that merges the string-encoded memory config protos from all hosts.final classReturns the result of a TPU compilation.final classAn op computes the size of the deduplication data from embedding core and returns the updated config.final classAn op computes tuple mask of deduplication data from embedding core.final classAn op that sets up the centralized structures for a distributed TPU system.final classSets up the centralized structures for a distributed TPU system.final classAn op that configures the TPUEmbedding software on a host.final classAn op that configures the TPUEmbedding software on a host.final classCrossReplicaSum<T extends TNumber>An Op to sum inputs across replicated TPU instances.final classAn op enabling differentiation of TPU Embeddings.final classAn op that executes the TPUEmbedding partitioner on the central configuration device and computes the HBM size (in bytes) required for TPUEmbedding operation.final classAn op returns the TPU task ID from TPU topology.final classThe GlobalIterId operationfinal classInfeedDequeue<T extends TType>A placeholder op for a value that will be fed into the computation.final classWhether TPU Embedding is initialized in a distributed TPU system.final classAn op merges elements of integer and float tensors into deduplication data as XLA tuple.final classA TPU core selector Op.final classOutfeedDequeue<T extends TType>Retrieves a single tensor from the computation outfeed.final classOutfeedDequeueV2<T extends TType>Retrieves a single tensor from the computation outfeed.final classPartitionedInput<T extends TType>An op that groups a list of partitioned inputs together.final classAn op which linearizes one Tensor value to an opaque variant tensor.final classAn op which linearizes multiple Tensor values to an opaque variant tensor.final classReplicatedInput<T extends TType>Connects N inputs to an N-way replicated TPU computation.final classRetrieve SGD embedding parameters.final classAn op that shuts down the TPU system.final classDeprecated.useCompilationResultinsteadfinal classDeprecated.useEmbeddingActivationsinsteadfinal classConverts XRT's uid handles to TensorFlow-friendly input format.final classTPUReplicatedInput<T extends TType>Deprecated.useReplicatedInputinsteadfinal classRound-robin load balancing on TPU cores.final classWorker heartbeat op.Classes in org.tensorflow.op.tpu that implement interfaces with type arguments of type OperandModifier and TypeClassDescriptionfinal classThe DTensorRestoreV2 operationfinal classOp that loads and executes a TPU program on a TPU device.final classOp that executes a program with optional in-place variable updates.final classFetches multiple values from infeed as an XLA tuple.final classRetrieve multiple values from the computation outfeed.final classRetrieve multiple values from the computation outfeed.final classCalls a function placed on a specified TPU device.final classPartitionedOutput<T extends TType>An op that demultiplexes a tensor to be sharded by XLA to a list of partitioned outputs outside the XLA computation.final classAn op that receives embedding activations on the TPU.final classReplicatedOutput<T extends TType>Connects N outputs from an N-way replicated TPU computation.final classThe TPUAnnotateTensorsWithDynamicShape operationfinal classOp that copies host tensor to device with dynamic shape support.final classTPUReplicatedOutput<T extends TType>Deprecated.useReplicatedOutputinsteadFields in org.tensorflow.op.tpu declared as OperandModifier and TypeFieldDescriptionLoadTPUEmbeddingAdadeltaParameters.Inputs.accumulatorsValue of accumulators used in the Adadelta optimization algorithm.LoadTPUEmbeddingAdagradMomentumParameters.Inputs.accumulatorsValue of accumulators used in the Adagrad Momentum optimization algorithm.LoadTPUEmbeddingAdagradParameters.Inputs.accumulatorsValue of accumulators used in the Adagrad optimization algorithm.LoadTPUEmbeddingFTRLParameters.Inputs.accumulatorsValue of accumulators used in the FTRL optimization algorithm.LoadTPUEmbeddingMDLAdagradLightParameters.Inputs.accumulatorsValue of accumulators used in the MDL Adagrad Light optimization algorithm.LoadTPUEmbeddingProximalAdagradParameters.Inputs.accumulatorsValue of accumulators used in the proximal Adagrad optimization algorithm.LoadTPUEmbeddingMDLAdagradLightParameters.Inputs.benefitsValue of benefits used in the MDL Adagrad Light optimization algorithm.GetMinibatchesInCsrWithPhysicalReplica.Inputs.colIdsThe colIds inputGetMinibatchSplitsWithPhysicalReplica.Inputs.colIdsThe colIds inputConfigureTPUEmbeddingHost.Inputs.commonConfigA string-encoded common configuration proto containing metadata about the TPUEmbedding partitioner output.ConfigureTPUEmbeddingMemory.Inputs.commonConfigA string-encoded CommonConfiguration proto containing metadata about the TPUEmbedding partitioner output and the HBM size (in bytes) required for operation.FinalizeTPUEmbedding.Inputs.commonConfigA string-encoded common configuration proto containing metadata about the TPUEmbedding partitioner output and the HBM size (in bytes) required for operation.CompileSucceededAssert.Inputs.compilationStatusThe compilationStatus inputDynamicEnqueueTPUEmbeddingArbitraryTensorBatch.Inputs.deviceOrdinalThe TPU device to use.DynamicEnqueueTPUEmbeddingRaggedTensorBatch.Inputs.deviceOrdinalThe deviceOrdinal inputOutfeedDequeueTupleV2.Inputs.deviceOrdinalAn int scalar tensor, representing the TPU device to use.OutfeedDequeueV2.Inputs.deviceOrdinalAn int scalar tensor, representing the TPU device to use.PartitionedCall.Inputs.deviceOrdinalThe TPU device ordinal to run the function on.EmbeddingActivations.Inputs.embeddingVariableA trainable variable, enabling optimizers to find this op.TPUEmbeddingActivations.Inputs.embeddingVariableA trainable variable, enabling optimizers to find this op.MergeDedupData.Inputs.floatTensorA 1-D float tensor, includes float elements of deduplication data tuple.TPUReshardVariables.Inputs.formatStateVarThe formatStateVar inputGetMinibatchesInCsrWithPhysicalReplica.Inputs.gainsThe gains inputGetMinibatchSplitsWithPhysicalReplica.Inputs.gainsThe gains inputAllToAll.Inputs.groupAssignmentAn int32 tensor with shape [num_groups, num_replicas_per_group].CrossReplicaSum.Inputs.groupAssignmentAn int32 tensor with shape [num_groups, num_replicas_per_group].GetMinibatchesInCsrWithPhysicalReplica.Inputs.idCountsThe idCounts inputConvertToCooTensor.Inputs.indicesOrRowSplitsThe indicesOrRowSplits inputAllToAll.Inputs.inputThe local input to the sum.CrossReplicaSum.Inputs.inputThe local input to the sum.InfeedEnqueue.Inputs.inputA tensor that will be provided using the infeed mechanism.InfeedEnqueuePrelinearizedBuffer.Inputs.inputA variant tensor representing linearized output.OutfeedEnqueue.Inputs.inputA tensor that will be inserted into the outfeed queue.Prelinearize.Inputs.inputA tensor that will be linearized.ReplicatedOutput.Inputs.inputThe input inputSplitDedupData.Inputs.inputAn XLA tuple including integer and float elements as deduplication data tuple.TPUReplicatedOutput.Inputs.inputThe input inputPartitionedOutput.Inputs.inputsA tensor which represents the full shape of partitioned tensors.MergeDedupData.Inputs.integerTensorA 1-D integer tensor, includes integer elements of deduplication data tuple.Execute.Inputs.keyThe key inputExecuteAndUpdateVariables.Inputs.keyThe key inputLoadTPUEmbeddingFrequencyEstimatorParameters.Inputs.lastHitStepValue of last_hit_step used in the frequency estimator optimization algorithm.LoadTPUEmbeddingFTRLParameters.Inputs.linearsValue of linears used in the FTRL optimization algorithm.LoadTPUEmbeddingProximalYogiParameters.Inputs.mThe m inputStoreMinibatchStatisticsInFdo.Inputs.maxIdsThe maxIds inputStoreMinibatchStatisticsInFdo.Inputs.maxUniquesThe maxUniques inputConfigureTPUEmbeddingHost.Inputs.memoryConfigA string-encoded memory config proto containing metadata about the memory allocations reserved for TPUEmbedding.FinalizeTPUEmbedding.Inputs.memoryConfigA string-encoded memory config proto containing metadata about the memory allocations reserved for TPUEmbedding.LoadTPUEmbeddingCenteredRMSPropParameters.Inputs.mgValue of mg used in the centered RMSProp optimization algorithm.DynamicEnqueueTPUEmbeddingArbitraryTensorBatch.Inputs.modeOverrideA string input that overrides the mode specified in the TPUEmbeddingConfiguration.DynamicEnqueueTPUEmbeddingRaggedTensorBatch.Inputs.modeOverrideThe modeOverride inputEnqueueTPUEmbeddingArbitraryTensorBatch.Inputs.modeOverrideA string input that overrides the mode specified in the TPUEmbeddingConfiguration.EnqueueTPUEmbeddingBatch.Inputs.modeOverrideA string input that overrides the mode specified in the TPUEmbeddingConfiguration.EnqueueTPUEmbeddingIntegerBatch.Inputs.modeOverrideA string input that overrides the mode specified in the TPUEmbeddingConfiguration.EnqueueTPUEmbeddingRaggedTensorBatch.Inputs.modeOverrideA string input that overrides the mode specified in the TPUEmbeddingConfiguration.EnqueueTPUEmbeddingSparseBatch.Inputs.modeOverrideA string input that overrides the mode specified in the TPUEmbeddingConfiguration.EnqueueTPUEmbeddingSparseTensorBatch.Inputs.modeOverrideA string input that overrides the mode specified in the TPUEmbeddingConfiguration.LoadTPUEmbeddingCenteredRMSPropParameters.Inputs.momValue of mom used in the centered RMSProp optimization algorithm.LoadTPUEmbeddingRMSPropParameters.Inputs.momValue of mom used in the RMSProp optimization algorithm.LoadTPUEmbeddingAdagradMomentumParameters.Inputs.momentaValue of momenta used in the Adagrad Momentum optimization algorithm.LoadTPUEmbeddingADAMParameters.Inputs.momentaValue of momenta used in the ADAM optimization algorithm.LoadTPUEmbeddingMomentumParameters.Inputs.momentaValue of momenta used in the Momentum optimization algorithm.LoadTPUEmbeddingCenteredRMSPropParameters.Inputs.msValue of ms used in the centered RMSProp optimization algorithm.LoadTPUEmbeddingRMSPropParameters.Inputs.msValue of ms used in the RMSProp optimization algorithm.TPUReshardVariables.Inputs.newFormatKeyThe newFormatKey inputLoadTPUEmbeddingAdadeltaParameters.Inputs.parametersValue of parameters used in the Adadelta optimization algorithm.LoadTPUEmbeddingAdagradMomentumParameters.Inputs.parametersValue of parameters used in the Adagrad Momentum optimization algorithm.LoadTPUEmbeddingAdagradParameters.Inputs.parametersValue of parameters used in the Adagrad optimization algorithm.LoadTPUEmbeddingADAMParameters.Inputs.parametersValue of parameters used in the ADAM optimization algorithm.LoadTPUEmbeddingCenteredRMSPropParameters.Inputs.parametersValue of parameters used in the centered RMSProp optimization algorithm.LoadTPUEmbeddingFrequencyEstimatorParameters.Inputs.parametersValue of parameters used in the frequency estimator optimization algorithm.LoadTPUEmbeddingFTRLParameters.Inputs.parametersValue of parameters used in the FTRL optimization algorithm.LoadTPUEmbeddingMDLAdagradLightParameters.Inputs.parametersValue of parameters used in the MDL Adagrad Light optimization algorithm.LoadTPUEmbeddingMomentumParameters.Inputs.parametersValue of parameters used in the Momentum optimization algorithm.LoadTPUEmbeddingProximalAdagradParameters.Inputs.parametersValue of parameters used in the proximal Adagrad optimization algorithm.LoadTPUEmbeddingProximalYogiParameters.Inputs.parametersThe parameters inputLoadTPUEmbeddingRMSPropParameters.Inputs.parametersValue of parameters used in the RMSProp optimization algorithm.LoadTPUEmbeddingStochasticGradientDescentParameters.Inputs.parametersValue of parameters used in the stochastic gradient descent optimization algorithm.DTensorRestore.Inputs.prefixThe prefix inputGetMinibatchesInCsrWithPhysicalReplica.Inputs.programKeyThe programKey inputGetMinibatchSplitsWithPhysicalReplica.Inputs.programKeyThe programKey inputStoreMinibatchStatisticsInFdo.Inputs.programKeyThe programKey inputWorkerHeartbeat.Inputs.requestA string tensor containing a serialized WorkerHeartbeatRequestGetMinibatchesInCsrWithPhysicalReplica.Inputs.rowIdsThe rowIds inputGetMinibatchSplitsWithPhysicalReplica.Inputs.rowIdsThe rowIds inputDTensorRestore.Inputs.shapeAndSlicesThe shapeAndSlices inputEmbeddingActivations.Inputs.slicedActivationsThe embedding activations Tensor to return.TPUEmbeddingActivations.Inputs.slicedActivationsThe embedding activations Tensor to return.GetMinibatchesInCsrWithPhysicalReplica.Inputs.splitsThe splits inputDTensorRestore.Inputs.tensorNamesThe tensorNames inputTpuHandleToProtoKey.Inputs.uidThe uid inputLoadTPUEmbeddingAdadeltaParameters.Inputs.updatesValue of updates used in the Adadelta optimization algorithm.LoadTPUEmbeddingProximalYogiParameters.Inputs.vThe v inputConvertToCooTensor.Inputs.valuesThe values inputLoadTPUEmbeddingADAMParameters.Inputs.velocitiesValue of velocities used in the ADAM optimization algorithm.ConvertToCooTensor.Inputs.weightsThe weights inputLoadTPUEmbeddingMDLAdagradLightParameters.Inputs.weightsValue of weights used in the MDL Adagrad Light optimization algorithm.Fields in org.tensorflow.op.tpu with type parameters of type OperandModifier and TypeFieldDescriptionDynamicEnqueueTPUEmbeddingArbitraryTensorBatch.Inputs.aggregationWeightsA list of rank 1 Tensors containing per training example aggregation weights.DynamicEnqueueTPUEmbeddingRaggedTensorBatch.Inputs.aggregationWeightsThe aggregationWeights inputEnqueueTPUEmbeddingArbitraryTensorBatch.Inputs.aggregationWeightsA list of rank 1 Tensors containing per training example aggregation weights.EnqueueTPUEmbeddingRaggedTensorBatch.Inputs.aggregationWeightsA list of rank 1 Tensors containing per training example aggregation weights.EnqueueTPUEmbeddingSparseBatch.Inputs.aggregationWeightsA list of rank 1 Tensors containing per sample -- i.e. per (training example, feature) -- aggregation weights.EnqueueTPUEmbeddingSparseTensorBatch.Inputs.aggregationWeightsA list of rank 1 Tensors containing per training example aggregation weights.Execute.Inputs.argsThe args inputExecuteAndUpdateVariables.Inputs.argsThe args inputPartitionedCall.Inputs.argsThe arguments to the function.LoadAllTPUEmbeddingParameters.Inputs.auxiliary1A list of tensors, one for each embedding table, containing the initial values of the first auxiliary optimization parameter to use in embedding training loop updates.LoadAllTPUEmbeddingParameters.Inputs.auxiliary2A list of tensors, one for each embedding table, containing the initial values of the second auxiliary optimization parameter to use in embedding training loop updates.LoadAllTPUEmbeddingParameters.Inputs.auxiliary3A list of tensors, one for each embedding table, containing the initial values of the third auxiliary optimization parameter to use in embedding training loop updates.LoadAllTPUEmbeddingParameters.Inputs.auxiliary4A list of tensors, one for each embedding table, containing the initial values of the second auxiliary optimization parameter to use in embedding training loop updates.LoadAllTPUEmbeddingParameters.Inputs.auxiliary5A list of tensors, one for each embedding table, containing the initial values of the third auxiliary optimization parameter to use in embedding training loop updates.LoadAllTPUEmbeddingParameters.Inputs.auxiliary6A list of tensors, one for each embedding table, containing the initial values of the second auxiliary optimization parameter to use in embedding training loop updates.LoadAllTPUEmbeddingParameters.Inputs.auxiliary7A list of tensors, one for each embedding table, containing the initial values of the third auxiliary optimization parameter to use in embedding training loop updates.EnqueueTPUEmbeddingBatch.Inputs.batchA list of 1D tensors, one for each embedding table, containing the batch inputs encoded as dist_belief.SparseFeatures protos.EnqueueTPUEmbeddingIntegerBatch.Inputs.batchA list of 1D tensors, one for each embedding table, containing the indices into the tables.Compile.Inputs.dynamicShapesThe dynamicShapes inputDynamicEnqueueTPUEmbeddingArbitraryTensorBatch.Inputs.embeddingIndicesA list of rank 1 Tensors, indices into the embedding tables.DynamicEnqueueTPUEmbeddingRaggedTensorBatch.Inputs.embeddingIndicesThe embeddingIndices inputEnqueueTPUEmbeddingArbitraryTensorBatch.Inputs.embeddingIndicesA list of rank 1 Tensors, indices into the embedding tables.EnqueueTPUEmbeddingRaggedTensorBatch.Inputs.embeddingIndicesA list of rank 1 Tensors, indices into the embedding tables.EnqueueTPUEmbeddingSparseBatch.Inputs.embeddingIndicesA list of rank 1 Tensors, indices into the embedding tables.EnqueueTPUEmbeddingSparseTensorBatch.Inputs.embeddingIndicesA list of rank 1 Tensors, indices into the embedding tables.Compile.Inputs.guaranteedConstantsThe guaranteedConstants inputInfeedEnqueueTuple.Inputs.inputsA list of tensors that will be provided using the infeed mechanism.OutfeedEnqueueTuple.Inputs.inputsA list of tensors that will be inserted into the outfeed queue as an XLA tuple.PartitionedInput.Inputs.inputsA list of partitioned inputs which must have the same shape.PrelinearizeTuple.Inputs.inputsA list of tensors that will be provided using the infeed mechanism.ReplicatedInput.Inputs.inputsThe inputs inputSendTPUEmbeddingGradients.Inputs.inputsA TensorList of gradients with which to update embedding tables.TPUReplicatedInput.Inputs.inputsThe inputs inputSendTPUEmbeddingGradients.Inputs.learningRatesA TensorList of float32 scalars, one for each dynamic learning rate tag: see the comments in //third_party/tensorflow/core/protobuf/tpu/optimization_parameters.proto.CollateTPUEmbeddingMemory.Inputs.memoryConfigsString-encoded memory config protos containing metadata about the memory allocations reserved for TPUEmbedding across all hosts.ConnectTPUEmbeddingHosts.Inputs.networkConfigsStrings containing metadata about the hostname and RPC port used for communication with all hosts.LoadAllTPUEmbeddingParameters.Inputs.parametersA list of tensors, one for each embedding table, containing the initial embedding table parameters to use in embedding lookups.EnqueueTPUEmbeddingSparseBatch.Inputs.sampleIndicesA list of rank 1 Tensors specifying the training example and feature to which the corresponding embedding_indices and aggregation_weights values belong. sample_indices[i] must equal b * nf + f, where nf is the number of features from the corresponding table, f is in [0, nf), and b is in [0, batch size).EnqueueTPUEmbeddingSparseTensorBatch.Inputs.sampleIndicesA list of rank 1 Tensors specifying the training example to which the corresponding embedding_indices and aggregation_weights values belong.DynamicEnqueueTPUEmbeddingArbitraryTensorBatch.Inputs.sampleIndicesOrRowSplitsA list of rank 2 Tensors specifying the training example to which the corresponding embedding_indices and aggregation_weights values belong.EnqueueTPUEmbeddingArbitraryTensorBatch.Inputs.sampleIndicesOrRowSplitsA list of rank 2 Tensors specifying the training example to which the corresponding embedding_indices and aggregation_weights values belong.DynamicEnqueueTPUEmbeddingRaggedTensorBatch.Inputs.sampleSplitsThe sampleSplits inputEnqueueTPUEmbeddingRaggedTensorBatch.Inputs.sampleSplitsA list of rank 1 Tensors specifying the break points for splitting embedding_indices and aggregation_weights into rows.TPUAnnotateTensorsWithDynamicShape.Inputs.tensorsThe tensors inputTPUCopyWithDynamicShape.Inputs.tensorsThe tensors inputUpdateTaskIdAndGlobalCoreArray.Inputs.tpuTaskIdToShardIdAn array of int32 that maps TPU task ID to shard ID.TPUCopyWithDynamicShape.Inputs.unpaddedSizesThe unpaddedSizes inputTPUReshardVariables.Inputs.varsThe vars inputMethods in org.tensorflow.op.tpu that return types with arguments of type OperandModifier and TypeMethodDescriptionDTensorRestore.iterator()Execute.iterator()ExecuteAndUpdateVariables.iterator()InfeedDequeueTuple.iterator()OutfeedDequeueTuple.iterator()OutfeedDequeueTupleV2.iterator()PartitionedCall.iterator()PartitionedOutput.iterator()RecvTPUEmbeddingActivations.iterator()ReplicatedOutput.iterator()TPUAnnotateTensorsWithDynamicShape.iterator()TPUCopyWithDynamicShape.iterator()TPUReplicatedOutput.iterator()Deprecated.Methods in org.tensorflow.op.tpu with parameters of type OperandModifier and TypeMethodDescriptionAllToAll.create(Scope scope, Operand<T> input, Operand<TInt32> groupAssignment, Long concatDimension, Long splitDimension, Long splitCount) Factory method to create a class wrapping a new AllToAll operation.static CompileSucceededAssertFactory method to create a class wrapping a new TPUCompileSucceededAssert operation.static ConfigureTPUEmbeddingHostConfigureTPUEmbeddingHost.create(Scope scope, Operand<TString> commonConfig, Operand<TString> memoryConfig, String config) Factory method to create a class wrapping a new ConfigureTPUEmbeddingHost operation.static ConfigureTPUEmbeddingMemoryFactory method to create a class wrapping a new ConfigureTPUEmbeddingMemory operation.static ConvertToCooTensorConvertToCooTensor.create(Scope scope, Operand<TInt32> indicesOrRowSplits, Operand<TInt32> values, Operand<TFloat32> weights, Long sampleCount, String combiner) Factory method to create a class wrapping a new ConvertToCooTensor operation.static <T extends TNumber>
CrossReplicaSum<T> Factory method to create a class wrapping a new CrossReplicaSum operation.static DTensorRestoreDTensorRestore.create(Scope scope, Operand<TString> prefix, Operand<TString> tensorNames, Operand<TString> shapeAndSlices, List<Shape> inputShapes, List<String> inputLayouts, List<Class<? extends TType>> dtypes) Factory method to create a class wrapping a new DTensorRestoreV2 operation.DynamicEnqueueTPUEmbeddingArbitraryTensorBatch.create(Scope scope, Iterable<Operand<? extends TNumber>> sampleIndicesOrRowSplits, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, Operand<TInt32> deviceOrdinal, DynamicEnqueueTPUEmbeddingArbitraryTensorBatch.Options... options) Factory method to create a class wrapping a new DynamicEnqueueTPUEmbeddingArbitraryTensorBatch operation.DynamicEnqueueTPUEmbeddingRaggedTensorBatch.create(Scope scope, Iterable<Operand<? extends TNumber>> sampleSplits, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, Operand<TInt32> deviceOrdinal, List<Long> tableIds, DynamicEnqueueTPUEmbeddingRaggedTensorBatch.Options... options) Factory method to create a class wrapping a new DynamicEnqueueTPUEmbeddingRaggedTensorBatch operation.static EmbeddingActivationsEmbeddingActivations.create(Scope scope, Operand<TFloat32> embeddingVariable, Operand<TFloat32> slicedActivations, Long tableId, Long lookupId) Factory method to create a class wrapping a new TPUEmbeddingActivations operation.EnqueueTPUEmbeddingArbitraryTensorBatch.create(Scope scope, Iterable<Operand<? extends TNumber>> sampleIndicesOrRowSplits, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, EnqueueTPUEmbeddingArbitraryTensorBatch.Options... options) Factory method to create a class wrapping a new EnqueueTPUEmbeddingArbitraryTensorBatch operation.static EnqueueTPUEmbeddingBatchEnqueueTPUEmbeddingBatch.create(Scope scope, Iterable<Operand<TString>> batch, Operand<TString> modeOverride, EnqueueTPUEmbeddingBatch.Options... options) Factory method to create a class wrapping a new EnqueueTPUEmbeddingBatch operation.EnqueueTPUEmbeddingIntegerBatch.create(Scope scope, Iterable<Operand<TInt32>> batch, Operand<TString> modeOverride, EnqueueTPUEmbeddingIntegerBatch.Options... options) Factory method to create a class wrapping a new EnqueueTPUEmbeddingIntegerBatch operation.EnqueueTPUEmbeddingRaggedTensorBatch.create(Scope scope, Iterable<Operand<? extends TNumber>> sampleSplits, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, List<Long> tableIds, EnqueueTPUEmbeddingRaggedTensorBatch.Options... options) Factory method to create a class wrapping a new EnqueueTPUEmbeddingRaggedTensorBatch operation.EnqueueTPUEmbeddingSparseBatch.create(Scope scope, Iterable<Operand<? extends TNumber>> sampleIndices, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, EnqueueTPUEmbeddingSparseBatch.Options... options) Factory method to create a class wrapping a new EnqueueTPUEmbeddingSparseBatch operation.EnqueueTPUEmbeddingSparseTensorBatch.create(Scope scope, Iterable<Operand<? extends TNumber>> sampleIndices, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, List<Long> tableIds, EnqueueTPUEmbeddingSparseTensorBatch.Options... options) Factory method to create a class wrapping a new EnqueueTPUEmbeddingSparseTensorBatch operation.static ExecuteExecute.create(Scope scope, Iterable<Operand<?>> args, Operand<TString> key, List<Class<? extends TType>> Tresults) Factory method to create a class wrapping a new TPUExecute operation.static ExecuteAndUpdateVariablesExecuteAndUpdateVariables.create(Scope scope, Iterable<Operand<?>> args, Operand<TString> key, List<Class<? extends TType>> Tresults, List<Long> deviceVarReadsIndices, List<Long> deviceVarUpdatesIndices) Factory method to create a class wrapping a new TPUExecuteAndUpdateVariables operation.static FinalizeTPUEmbeddingFinalizeTPUEmbedding.create(Scope scope, Operand<TString> commonConfig, Operand<TString> memoryConfig) Factory method to create a class wrapping a new FinalizeTPUEmbeddingV2 operation.GetMinibatchesInCsrWithPhysicalReplica.create(Scope scope, Operand<TString> programKey, Operand<TInt32> rowIds, Operand<TInt32> colIds, Operand<TFloat32> gains, Operand<TInt64> splits, Operand<TInt32> idCounts, Long sampleCount, Long numReplica, Long maxMinibatchesPerSc, Long maxIdsPerChipPerSample, Long tableVocabSize, Long featureWidth, Long numScPerChip, String tableName, String miniBatchInCsr) Factory method to create a class wrapping a new GetMinibatchesInCsrWithPhysicalReplica operation.GetMinibatchSplitsWithPhysicalReplica.create(Scope scope, Operand<TString> programKey, Operand<TInt32> rowIds, Operand<TInt32> colIds, Operand<TFloat32> gains, Long sampleCount, Long numReplica, Long tableVocabSize, Long featureWidth, Long numScPerChip, String tableName, String miniBatchSplits) Factory method to create a class wrapping a new GetMinibatchSplitsWithPhysicalReplica operation.static InfeedEnqueueInfeedEnqueue.create(Scope scope, Operand<? extends TType> input, InfeedEnqueue.Options... options) Factory method to create a class wrapping a new InfeedEnqueue operation.InfeedEnqueuePrelinearizedBuffer.create(Scope scope, Operand<? extends TType> input, InfeedEnqueuePrelinearizedBuffer.Options... options) Factory method to create a class wrapping a new InfeedEnqueuePrelinearizedBuffer operation.LoadTPUEmbeddingAdadeltaParameters.create(Scope scope, Operand<TFloat32> parameters, Operand<TFloat32> accumulators, Operand<TFloat32> updates, Long numShards, Long shardId, LoadTPUEmbeddingAdadeltaParameters.Options... options) Factory method to create a class wrapping a new LoadTPUEmbeddingAdadeltaParameters operation.LoadTPUEmbeddingAdagradMomentumParameters.create(Scope scope, Operand<TFloat32> parameters, Operand<TFloat32> accumulators, Operand<TFloat32> momenta, Long numShards, Long shardId, LoadTPUEmbeddingAdagradMomentumParameters.Options... options) Factory method to create a class wrapping a new LoadTPUEmbeddingAdagradMomentumParameters operation.LoadTPUEmbeddingAdagradParameters.create(Scope scope, Operand<TFloat32> parameters, Operand<TFloat32> accumulators, Long numShards, Long shardId, LoadTPUEmbeddingAdagradParameters.Options... options) Factory method to create a class wrapping a new LoadTPUEmbeddingAdagradParameters operation.LoadTPUEmbeddingADAMParameters.create(Scope scope, Operand<TFloat32> parameters, Operand<TFloat32> momenta, Operand<TFloat32> velocities, Long numShards, Long shardId, LoadTPUEmbeddingADAMParameters.Options... options) Factory method to create a class wrapping a new LoadTPUEmbeddingADAMParameters operation.LoadTPUEmbeddingCenteredRMSPropParameters.create(Scope scope, Operand<TFloat32> parameters, Operand<TFloat32> ms, Operand<TFloat32> mom, Operand<TFloat32> mg, Long numShards, Long shardId, LoadTPUEmbeddingCenteredRMSPropParameters.Options... options) Factory method to create a class wrapping a new LoadTPUEmbeddingCenteredRMSPropParameters operation.LoadTPUEmbeddingFrequencyEstimatorParameters.create(Scope scope, Operand<TFloat32> parameters, Operand<TFloat32> lastHitStep, Long numShards, Long shardId, LoadTPUEmbeddingFrequencyEstimatorParameters.Options... options) Factory method to create a class wrapping a new LoadTPUEmbeddingFrequencyEstimatorParameters operation.LoadTPUEmbeddingFTRLParameters.create(Scope scope, Operand<TFloat32> parameters, Operand<TFloat32> accumulators, Operand<TFloat32> linears, Long numShards, Long shardId, LoadTPUEmbeddingFTRLParameters.Options... options) Factory method to create a class wrapping a new LoadTPUEmbeddingFTRLParameters operation.LoadTPUEmbeddingMDLAdagradLightParameters.create(Scope scope, Operand<TFloat32> parameters, Operand<TFloat32> accumulators, Operand<TFloat32> weights, Operand<TFloat32> benefits, Long numShards, Long shardId, LoadTPUEmbeddingMDLAdagradLightParameters.Options... options) Factory method to create a class wrapping a new LoadTPUEmbeddingMDLAdagradLightParameters operation.LoadTPUEmbeddingMomentumParameters.create(Scope scope, Operand<TFloat32> parameters, Operand<TFloat32> momenta, Long numShards, Long shardId, LoadTPUEmbeddingMomentumParameters.Options... options) Factory method to create a class wrapping a new LoadTPUEmbeddingMomentumParameters operation.LoadTPUEmbeddingProximalAdagradParameters.create(Scope scope, Operand<TFloat32> parameters, Operand<TFloat32> accumulators, Long numShards, Long shardId, LoadTPUEmbeddingProximalAdagradParameters.Options... options) Factory method to create a class wrapping a new LoadTPUEmbeddingProximalAdagradParameters operation.LoadTPUEmbeddingProximalYogiParameters.create(Scope scope, Operand<TFloat32> parameters, Operand<TFloat32> v, Operand<TFloat32> m, Long numShards, Long shardId, LoadTPUEmbeddingProximalYogiParameters.Options... options) Factory method to create a class wrapping a new LoadTPUEmbeddingProximalYogiParameters operation.LoadTPUEmbeddingRMSPropParameters.create(Scope scope, Operand<TFloat32> parameters, Operand<TFloat32> ms, Operand<TFloat32> mom, Long numShards, Long shardId, LoadTPUEmbeddingRMSPropParameters.Options... options) Factory method to create a class wrapping a new LoadTPUEmbeddingRMSPropParameters operation.LoadTPUEmbeddingStochasticGradientDescentParameters.create(Scope scope, Operand<TFloat32> parameters, Long numShards, Long shardId, LoadTPUEmbeddingStochasticGradientDescentParameters.Options... options) Factory method to create a class wrapping a new LoadTPUEmbeddingStochasticGradientDescentParameters operation.static MergeDedupDataMergeDedupData.create(Scope scope, Operand<? extends TNumber> integerTensor, Operand<? extends TNumber> floatTensor, String tupleMask, MergeDedupData.Options... options) Factory method to create a class wrapping a new MergeDedupData operation.static OutfeedDequeueTupleV2OutfeedDequeueTupleV2.create(Scope scope, Operand<TInt32> deviceOrdinal, List<Class<? extends TType>> dtypes, List<Shape> shapes) Factory method to create a class wrapping a new OutfeedDequeueTupleV2 operation.static <T extends TType>
OutfeedDequeueV2<T> Factory method to create a class wrapping a new OutfeedDequeueV2 operation.static OutfeedEnqueueFactory method to create a class wrapping a new OutfeedEnqueue operation.static PartitionedCallPartitionedCall.create(Scope scope, Iterable<Operand<?>> args, Operand<TInt32> deviceOrdinal, List<Class<? extends TType>> Tout, ConcreteFunction f, PartitionedCall.Options... options) Factory method to create a class wrapping a new TPUPartitionedCall operation.static <T extends TType>
PartitionedOutput<T> Factory method to create a class wrapping a new TPUPartitionedOutputV2 operation.static PrelinearizePrelinearize.create(Scope scope, Operand<? extends TType> input, Prelinearize.Options... options) Factory method to create a class wrapping a new Prelinearize operation.static <T extends TType>
ReplicatedOutput<T> Factory method to create a class wrapping a new TPUReplicatedOutput operation.static <T extends TNumber, U extends TNumber>
SplitDedupData<T, U> SplitDedupData.create(Scope scope, Operand<? extends TType> input, Class<T> integerType, Class<U> floatType, String tupleMask, SplitDedupData.Options... options) Factory method to create a class wrapping a new SplitDedupData operation.StoreMinibatchStatisticsInFdo.create(Scope scope, Operand<TString> programKey, Operand<TInt32> maxIds, Operand<TInt32> maxUniques, Long sampleCount, Long numReplica, Long featureWidth, Long numScPerChip, String tableName, String miniBatchSplits) Factory method to create a class wrapping a new StoreMinibatchStatisticsInFdo operation.static TPUEmbeddingActivationsTPUEmbeddingActivations.create(Scope scope, Operand<TFloat32> embeddingVariable, Operand<TFloat32> slicedActivations, Long tableId, Long lookupId) Deprecated.Factory method to create a class wrapping a new TPUEmbeddingActivations operation.static TpuHandleToProtoKeyFactory method to create a class wrapping a new TpuHandleToProtoKey operation.static <T extends TType>
TPUReplicatedOutput<T> Deprecated.Factory method to create a class wrapping a new TPUReplicatedOutput operation.static TPUReshardVariablesTPUReshardVariables.create(Scope scope, Iterable<Operand<? extends TType>> vars, Operand<TString> newFormatKey, Operand<? extends TType> formatStateVar) Factory method to create a class wrapping a new TPUReshardVariables operation.static WorkerHeartbeatFactory method to create a class wrapping a new WorkerHeartbeat operation.Method parameters in org.tensorflow.op.tpu with type arguments of type OperandModifier and TypeMethodDescriptionstatic CollateTPUEmbeddingMemoryFactory method to create a class wrapping a new CollateTPUEmbeddingMemory operation.static CompileCompile.create(Scope scope, Iterable<Operand<TInt64>> dynamicShapes, Iterable<Operand<?>> guaranteedConstants, Long numComputations, ConcreteFunction function, String metadata) Factory method to create a class wrapping a new TPUCompile operation.static ConnectTPUEmbeddingHostsFactory method to create a class wrapping a new ConnectTPUEmbeddingHosts operation.DynamicEnqueueTPUEmbeddingArbitraryTensorBatch.create(Scope scope, Iterable<Operand<? extends TNumber>> sampleIndicesOrRowSplits, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, Operand<TInt32> deviceOrdinal, DynamicEnqueueTPUEmbeddingArbitraryTensorBatch.Options... options) Factory method to create a class wrapping a new DynamicEnqueueTPUEmbeddingArbitraryTensorBatch operation.DynamicEnqueueTPUEmbeddingRaggedTensorBatch.create(Scope scope, Iterable<Operand<? extends TNumber>> sampleSplits, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, Operand<TInt32> deviceOrdinal, List<Long> tableIds, DynamicEnqueueTPUEmbeddingRaggedTensorBatch.Options... options) Factory method to create a class wrapping a new DynamicEnqueueTPUEmbeddingRaggedTensorBatch operation.EnqueueTPUEmbeddingArbitraryTensorBatch.create(Scope scope, Iterable<Operand<? extends TNumber>> sampleIndicesOrRowSplits, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, EnqueueTPUEmbeddingArbitraryTensorBatch.Options... options) Factory method to create a class wrapping a new EnqueueTPUEmbeddingArbitraryTensorBatch operation.static EnqueueTPUEmbeddingBatchEnqueueTPUEmbeddingBatch.create(Scope scope, Iterable<Operand<TString>> batch, Operand<TString> modeOverride, EnqueueTPUEmbeddingBatch.Options... options) Factory method to create a class wrapping a new EnqueueTPUEmbeddingBatch operation.EnqueueTPUEmbeddingIntegerBatch.create(Scope scope, Iterable<Operand<TInt32>> batch, Operand<TString> modeOverride, EnqueueTPUEmbeddingIntegerBatch.Options... options) Factory method to create a class wrapping a new EnqueueTPUEmbeddingIntegerBatch operation.EnqueueTPUEmbeddingRaggedTensorBatch.create(Scope scope, Iterable<Operand<? extends TNumber>> sampleSplits, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, List<Long> tableIds, EnqueueTPUEmbeddingRaggedTensorBatch.Options... options) Factory method to create a class wrapping a new EnqueueTPUEmbeddingRaggedTensorBatch operation.EnqueueTPUEmbeddingSparseBatch.create(Scope scope, Iterable<Operand<? extends TNumber>> sampleIndices, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, EnqueueTPUEmbeddingSparseBatch.Options... options) Factory method to create a class wrapping a new EnqueueTPUEmbeddingSparseBatch operation.EnqueueTPUEmbeddingSparseTensorBatch.create(Scope scope, Iterable<Operand<? extends TNumber>> sampleIndices, Iterable<Operand<? extends TNumber>> embeddingIndices, Iterable<Operand<? extends TNumber>> aggregationWeights, Operand<TString> modeOverride, List<Long> tableIds, EnqueueTPUEmbeddingSparseTensorBatch.Options... options) Factory method to create a class wrapping a new EnqueueTPUEmbeddingSparseTensorBatch operation.static ExecuteExecute.create(Scope scope, Iterable<Operand<?>> args, Operand<TString> key, List<Class<? extends TType>> Tresults) Factory method to create a class wrapping a new TPUExecute operation.static ExecuteAndUpdateVariablesExecuteAndUpdateVariables.create(Scope scope, Iterable<Operand<?>> args, Operand<TString> key, List<Class<? extends TType>> Tresults, List<Long> deviceVarReadsIndices, List<Long> deviceVarUpdatesIndices) Factory method to create a class wrapping a new TPUExecuteAndUpdateVariables operation.static InfeedEnqueueTupleInfeedEnqueueTuple.create(Scope scope, Iterable<Operand<?>> inputs, List<Shape> shapes, InfeedEnqueueTuple.Options... options) Factory method to create a class wrapping a new InfeedEnqueueTuple operation.LoadAllTPUEmbeddingParameters.create(Scope scope, Iterable<Operand<TFloat32>> parameters, Iterable<Operand<TFloat32>> auxiliary1, Iterable<Operand<TFloat32>> auxiliary2, Iterable<Operand<TFloat32>> auxiliary3, Iterable<Operand<TFloat32>> auxiliary4, Iterable<Operand<TFloat32>> auxiliary5, Iterable<Operand<TFloat32>> auxiliary6, Iterable<Operand<TFloat32>> auxiliary7, String config, Long numShards, Long shardId) Factory method to create a class wrapping a new LoadAllTPUEmbeddingParameters operation.static OutfeedEnqueueTupleFactory method to create a class wrapping a new OutfeedEnqueueTuple operation.static PartitionedCallPartitionedCall.create(Scope scope, Iterable<Operand<?>> args, Operand<TInt32> deviceOrdinal, List<Class<? extends TType>> Tout, ConcreteFunction f, PartitionedCall.Options... options) Factory method to create a class wrapping a new TPUPartitionedCall operation.static <T extends TType>
PartitionedInput<T> PartitionedInput.create(Scope scope, Iterable<Operand<T>> inputs, List<Long> partitionDims, PartitionedInput.Options... options) Factory method to create a class wrapping a new TPUPartitionedInputV2 operation.static PrelinearizeTuplePrelinearizeTuple.create(Scope scope, Iterable<Operand<?>> inputs, List<Shape> shapes, PrelinearizeTuple.Options... options) Factory method to create a class wrapping a new PrelinearizeTuple operation.static <T extends TType>
ReplicatedInput<T> ReplicatedInput.create(Scope scope, Iterable<Operand<T>> inputs, ReplicatedInput.Options... options) Factory method to create a class wrapping a new TPUReplicatedInput operation.static SendTPUEmbeddingGradientsSendTPUEmbeddingGradients.create(Scope scope, Iterable<Operand<TFloat32>> inputs, Iterable<Operand<TFloat32>> learningRates, String config, SendTPUEmbeddingGradients.Options... options) Factory method to create a class wrapping a new SendTPUEmbeddingGradients operation.Factory method to create a class wrapping a new TPUAnnotateTensorsWithDynamicShape operation.static TPUCopyWithDynamicShapeTPUCopyWithDynamicShape.create(Scope scope, Iterable<Operand<?>> tensors, Iterable<Operand<TInt32>> unpaddedSizes) Factory method to create a class wrapping a new TPUCopyWithDynamicShape operation.static <T extends TType>
TPUReplicatedInput<T> TPUReplicatedInput.create(Scope scope, Iterable<Operand<T>> inputs, TPUReplicatedInput.Options... options) Deprecated.Factory method to create a class wrapping a new TPUReplicatedInput operation.static TPUReshardVariablesTPUReshardVariables.create(Scope scope, Iterable<Operand<? extends TType>> vars, Operand<TString> newFormatKey, Operand<? extends TType> formatStateVar) Factory method to create a class wrapping a new TPUReshardVariables operation.Factory method to create a class wrapping a new UpdateTaskIdAndGlobalCoreArray operation. -
Uses of Operand in org.tensorflow.op.train
Classes in org.tensorflow.op.train that implement OperandModifier and TypeClassDescriptionfinal classReturns the number of gradients aggregated in the given accumulators.final classAccumulatorTakeGradient<T extends TType>Extracts the average gradient in the given ConditionalAccumulator.final classApplyAdadelta<T extends TType>Update '*var' according to the adadelta scheme. accum = rho() * accum + (1 - rho()) * grad.square(); update = (update_accum + epsilon).sqrt() * (accum + epsilon()).rsqrt() * grad; update_accum = rho() * update_accum + (1 - rho()) * update.square(); var -= update;final classApplyAdagrad<T extends TType>Update '*var' according to the adagrad scheme. accum += grad * grad var -= lr * grad * (1 / sqrt(accum))final classApplyAdagradDa<T extends TType>Update '*var' according to the proximal adagrad scheme.final classApplyAdagradV2<T extends TType>Update '*var' according to the adagrad scheme. accum += grad * grad var -= lr * grad * (1 / sqrt(accum))final classUpdate '*var' according to the Adam algorithm.final classApplyAdaMax<T extends TType>Update '*var' according to the AdaMax algorithm. m_t <- beta1 * m_{t-1} + (1 - beta1) * g v_t <- max(beta2 * v_{t-1}, abs(g)) variable <- variable - learning_rate / (1 - beta1^t) * m_t / (v_t + epsilon)final classApplyAddSign<T extends TType>Update '*var' according to the AddSign update. m_t <- beta1 * m_{t-1} + (1 - beta1) * g update <- (alpha + sign_decay * sign(g) *sign(m)) * g variable <- variable - lr_t * updatefinal classApplyCenteredRmsProp<T extends TType>Update '*var' according to the centered RMSProp algorithm.final classUpdate '*var' according to the Ftrl-proximal scheme. grad_with_shrinkage = grad + 2 * l2_shrinkage * var accum_new = accum + grad * grad linear += grad_with_shrinkage - (accum_new^(-lr_power) - accum^(-lr_power)) / lr * var quadratic = 1.0 / (accum_new^(lr_power) * lr) + 2 * l2 var = (sign(linear) * l1 - linear) / quadratic if |linear| > l1 else 0.0 accum = accum_newfinal classApplyGradientDescent<T extends TType>Update '*var' by subtracting 'alpha' * 'delta' from it.final classApplyMomentum<T extends TType>Update '*var' according to the momentum scheme.final classApplyPowerSign<T extends TType>Update '*var' according to the AddSign update. m_t <- beta1 * m_{t-1} + (1 - beta1) * g update <- exp(logbase * sign_decay * sign(g) * sign(m_t)) * g variable <- variable - lr_t * updatefinal classApplyProximalAdagrad<T extends TType>Update '*var' and '*accum' according to FOBOS with Adagrad learning rate. accum += grad * grad prox_v = var - lr * grad * (1 / sqrt(accum)) var = sign(prox_v)/(1+lrl2) * max{|prox_v|-lrl1,0}final classApplyProximalGradientDescent<T extends TType>Update '*var' as FOBOS algorithm with fixed learning rate. prox_v = var - alpha * delta var = sign(prox_v)/(1+alphal2) * max{|prox_v|-alphal1,0}final classApplyRmsProp<T extends TType>Update '*var' according to the RMSProp algorithm.final classBatchMatMul<V extends TType>Multiplies slices of two tensors in batches.final classComputes the static batch size of a dataset sans partial batches.final classA conditional accumulator for aggregating gradients.final classPreventGradient<T extends TType>An identity op that triggers an error if a gradient is requested.final classReturns the number of gradients aggregated in the given accumulators.final classResourceAccumulatorTakeGradient<T extends TType>Extracts the average gradient in the given ConditionalAccumulator.final classA conditional accumulator for aggregating gradients.final classRestoreSlice<T extends TType>Restores a tensor from checkpoint files.final classComputes fingerprints of the input strings.final classSparseApplyAdadelta<T extends TType>var: Should be from a Variable().final classSparseApplyAdagrad<T extends TType>Update relevant entries in '*var' and '*accum' according to the adagrad scheme.final classSparseApplyAdagradDa<T extends TType>Update entries in '*var' and '*accum' according to the proximal adagrad scheme.final classSparseApplyCenteredRmsProp<T extends TType>Update '*var' according to the centered RMSProp algorithm.final classSparseApplyFtrl<T extends TType>Update relevant entries in '*var' according to the Ftrl-proximal scheme.final classSparseApplyMomentum<T extends TType>Update relevant entries in '*var' and '*accum' according to the momentum scheme.final classSparseApplyProximalAdagrad<T extends TType>Sparse update entries in '*var' and '*accum' according to FOBOS algorithm.final classSparseApplyProximalGradientDescent<T extends TType>Sparse update '*var' as FOBOS algorithm with fixed learning rate.final classSparseApplyRmsProp<T extends TType>Update '*var' according to the RMSProp algorithm.final classReturns the gradient ofTile.Classes in org.tensorflow.op.train that implement interfaces with type arguments of type OperandModifier and TypeClassDescriptionfinal classRestores tensors from a V2 checkpoint.final classComputes the gradient function for function f via backpropagation.Fields in org.tensorflow.op.train declared as OperandModifier and TypeFieldDescriptionApplyAdadelta.Inputs.accumShould be from a Variable().ApplyAdagrad.Inputs.accumShould be from a Variable().ApplyAdagradV2.Inputs.accumShould be from a Variable().ApplyFtrl.Inputs.accumShould be from a Variable().ApplyMomentum.Inputs.accumShould be from a Variable().ApplyProximalAdagrad.Inputs.accumShould be from a Variable().ResourceApplyAdadelta.Inputs.accumShould be from a Variable().ResourceApplyAdagrad.Inputs.accumShould be from a Variable().ResourceApplyFtrl.Inputs.accumShould be from a Variable().ResourceApplyKerasMomentum.Inputs.accumShould be from a Variable().ResourceApplyMomentum.Inputs.accumShould be from a Variable().ResourceApplyProximalAdagrad.Inputs.accumShould be from a Variable().ResourceSparseApplyAdadelta.Inputs.accumShould be from a Variable().ResourceSparseApplyAdagrad.Inputs.accumShould be from a Variable().ResourceSparseApplyAdagradV2.Inputs.accumShould be from a Variable().ResourceSparseApplyFtrl.Inputs.accumShould be from a Variable().ResourceSparseApplyKerasMomentum.Inputs.accumShould be from a Variable().ResourceSparseApplyMomentum.Inputs.accumShould be from a Variable().ResourceSparseApplyProximalAdagrad.Inputs.accumShould be from a Variable().SparseApplyAdadelta.Inputs.accumShould be from a Variable().SparseApplyAdagrad.Inputs.accumShould be from a Variable().SparseApplyFtrl.Inputs.accumShould be from a Variable().SparseApplyMomentum.Inputs.accumShould be from a Variable().SparseApplyProximalAdagrad.Inputs.accumShould be from a Variable().ApplyAdadelta.Inputs.accumUpdateShould be from a Variable().ResourceApplyAdadelta.Inputs.accumUpdateShould be from a Variable().ResourceSparseApplyAdadelta.Inputs.accumUpdate: Should be from a Variable().SparseApplyAdadelta.Inputs.accumUpdate: Should be from a Variable().DistributedSave.Inputs.addressThe address inputApplyAddSign.Inputs.alphaMust be a scalar.ApplyGradientDescent.Inputs.alphaScaling factor.ApplyProximalGradientDescent.Inputs.alphaScaling factor.ResourceApplyAddSign.Inputs.alphaMust be a scalar.ResourceApplyGradientDescent.Inputs.alphaScaling factor.ResourceApplyProximalGradientDescent.Inputs.alphaScaling factor.ResourceSparseApplyProximalGradientDescent.Inputs.alphaScaling factor.SparseApplyProximalGradientDescent.Inputs.alphaScaling factor.ApplyAddSign.Inputs.betaMust be a scalar.ApplyPowerSign.Inputs.betaMust be a scalar.ResourceApplyAddSign.Inputs.betaMust be a scalar.ResourceApplyPowerSign.Inputs.betaMust be a scalar.ApplyAdam.Inputs.beta1Momentum factor.ApplyAdaMax.Inputs.beta1Momentum factor.ResourceApplyAdam.Inputs.beta1Momentum factor.ResourceApplyAdaMax.Inputs.beta1Momentum factor.ResourceApplyAdamWithAmsgrad.Inputs.beta1Momentum factor.ApplyAdam.Inputs.beta1PowerMust be a scalar.ApplyAdaMax.Inputs.beta1PowerMust be a scalar.ResourceApplyAdam.Inputs.beta1PowerMust be a scalar.ResourceApplyAdaMax.Inputs.beta1PowerMust be a scalar.ResourceApplyAdamWithAmsgrad.Inputs.beta1PowerMust be a scalar.ApplyAdam.Inputs.beta2Momentum factor.ApplyAdaMax.Inputs.beta2Momentum factor.ResourceApplyAdam.Inputs.beta2Momentum factor.ResourceApplyAdaMax.Inputs.beta2Momentum factor.ResourceApplyAdamWithAmsgrad.Inputs.beta2Momentum factor.ApplyAdam.Inputs.beta2PowerMust be a scalar.ResourceApplyAdam.Inputs.beta2PowerMust be a scalar.ResourceApplyAdamWithAmsgrad.Inputs.beta2PowerMust be a scalar.MergeV2Checkpoints.Inputs.checkpointPrefixesprefixes of V2 checkpoints to merge.DistributedSave.Inputs.datasetThe dataset inputApplyGradientDescent.Inputs.deltaThe change.ApplyProximalGradientDescent.Inputs.deltaThe change.ResourceApplyGradientDescent.Inputs.deltaThe change.ResourceApplyProximalGradientDescent.Inputs.deltaThe change.MergeV2Checkpoints.Inputs.destinationPrefixscalar.DistributedSave.Inputs.directoryThe directory inputApplyAdadelta.Inputs.epsilonConstant factor.ApplyAdagradV2.Inputs.epsilonConstant factor.ApplyAdam.Inputs.epsilonRidge term.ApplyAdaMax.Inputs.epsilonRidge term.ApplyCenteredRmsProp.Inputs.epsilonRidge term.ApplyRmsProp.Inputs.epsilonRidge term.ResourceApplyAdadelta.Inputs.epsilonConstant factor.ResourceApplyAdagrad.Inputs.epsilonConstant factor.ResourceApplyAdam.Inputs.epsilonRidge term.ResourceApplyAdaMax.Inputs.epsilonRidge term.ResourceApplyAdamWithAmsgrad.Inputs.epsilonRidge term.ResourceApplyCenteredRmsProp.Inputs.epsilonRidge term.ResourceApplyRmsProp.Inputs.epsilonRidge term.ResourceSparseApplyAdadelta.Inputs.epsilonConstant factor.ResourceSparseApplyAdagradV2.Inputs.epsilonConstant factor.ResourceSparseApplyCenteredRmsProp.Inputs.epsilonRidge term.ResourceSparseApplyRmsProp.Inputs.epsilonRidge term.SparseApplyAdadelta.Inputs.epsilonConstant factor.SparseApplyAdagrad.Inputs.epsilonConstant factor.SparseApplyCenteredRmsProp.Inputs.epsilonRidge term.SparseApplyRmsProp.Inputs.epsilonRidge term.SdcaOptimizer.Inputs.exampleLabelsa vector which contains the label/target associated with each example.NegTrain.Inputs.examplesA vector of word ids.SdcaOptimizer.Inputs.exampleStateDataa list of vectors containing the example state data.SdcaOptimizer.Inputs.exampleWeightsa vector which contains the weight associated with each example.SaveSlices.Inputs.filenameMust have a single element.RestoreSlice.Inputs.filePatternMust have a single element.ApplyAdagradDa.Inputs.globalStepTraining step number.ResourceApplyAdagradDa.Inputs.globalStepTraining step number.ResourceSparseApplyAdagradDa.Inputs.globalStepTraining step number.SparseApplyAdagradDa.Inputs.globalStepTraining step number.ApplyAdadelta.Inputs.gradThe gradient.ApplyAdagrad.Inputs.gradThe gradient.ApplyAdagradDa.Inputs.gradThe gradient.ApplyAdagradV2.Inputs.gradThe gradient.ApplyAdam.Inputs.gradThe gradient.ApplyAdaMax.Inputs.gradThe gradient.ApplyAddSign.Inputs.gradThe gradient.ApplyCenteredRmsProp.Inputs.gradThe gradient.ApplyFtrl.Inputs.gradThe gradient.ApplyMomentum.Inputs.gradThe gradient.ApplyPowerSign.Inputs.gradThe gradient.ApplyProximalAdagrad.Inputs.gradThe gradient.ApplyRmsProp.Inputs.gradThe gradient.ResourceApplyAdadelta.Inputs.gradThe gradient.ResourceApplyAdagrad.Inputs.gradThe gradient.ResourceApplyAdagradDa.Inputs.gradThe gradient.ResourceApplyAdam.Inputs.gradThe gradient.ResourceApplyAdaMax.Inputs.gradThe gradient.ResourceApplyAdamWithAmsgrad.Inputs.gradThe gradient.ResourceApplyAddSign.Inputs.gradThe gradient.ResourceApplyCenteredRmsProp.Inputs.gradThe gradient.ResourceApplyFtrl.Inputs.gradThe gradient.ResourceApplyKerasMomentum.Inputs.gradThe gradient.ResourceApplyMomentum.Inputs.gradThe gradient.ResourceApplyPowerSign.Inputs.gradThe gradient.ResourceApplyProximalAdagrad.Inputs.gradThe gradient.ResourceApplyRmsProp.Inputs.gradThe gradient.ResourceSparseApplyAdadelta.Inputs.gradThe gradient.ResourceSparseApplyAdagrad.Inputs.gradThe gradient.ResourceSparseApplyAdagradDa.Inputs.gradThe gradient.ResourceSparseApplyAdagradV2.Inputs.gradThe gradient.ResourceSparseApplyCenteredRmsProp.Inputs.gradThe gradient.ResourceSparseApplyFtrl.Inputs.gradThe gradient.ResourceSparseApplyKerasMomentum.Inputs.gradThe gradient.ResourceSparseApplyMomentum.Inputs.gradThe gradient.ResourceSparseApplyProximalAdagrad.Inputs.gradThe gradient.ResourceSparseApplyProximalGradientDescent.Inputs.gradThe gradient.ResourceSparseApplyRmsProp.Inputs.gradThe gradient.SparseApplyAdadelta.Inputs.gradThe gradient.SparseApplyAdagrad.Inputs.gradThe gradient.SparseApplyAdagradDa.Inputs.gradThe gradient.SparseApplyCenteredRmsProp.Inputs.gradThe gradient.SparseApplyFtrl.Inputs.gradThe gradient.SparseApplyMomentum.Inputs.gradThe gradient.SparseApplyProximalAdagrad.Inputs.gradThe gradient.SparseApplyProximalGradientDescent.Inputs.gradThe gradient.SparseApplyRmsProp.Inputs.gradThe gradient.AccumulatorApplyGradient.Inputs.gradientA tensor of the gradient to be accumulated.ResourceAccumulatorApplyGradient.Inputs.gradientA tensor of the gradient to be accumulated.ApplyAdagradDa.Inputs.gradientAccumulatorShould be from a Variable().ResourceApplyAdagradDa.Inputs.gradientAccumulatorShould be from a Variable().ResourceSparseApplyAdagradDa.Inputs.gradientAccumulatorShould be from a Variable().SparseApplyAdagradDa.Inputs.gradientAccumulatorShould be from a Variable().ApplyAdagradDa.Inputs.gradientSquaredAccumulatorShould be from a Variable().ResourceApplyAdagradDa.Inputs.gradientSquaredAccumulatorShould be from a Variable().ResourceSparseApplyAdagradDa.Inputs.gradientSquaredAccumulatorShould be from a Variable().SparseApplyAdagradDa.Inputs.gradientSquaredAccumulatorShould be from a Variable().AccumulatorApplyGradient.Inputs.handleThe handle to a accumulator.AccumulatorNumAccumulated.Inputs.handleThe handle to an accumulator.AccumulatorSetGlobalStep.Inputs.handleThe handle to an accumulator.AccumulatorTakeGradient.Inputs.handleThe handle to an accumulator.ResourceAccumulatorApplyGradient.Inputs.handleThe handle to a accumulator.ResourceAccumulatorNumAccumulated.Inputs.handleThe handle to an accumulator.ResourceAccumulatorSetGlobalStep.Inputs.handleThe handle to an accumulator.ResourceAccumulatorTakeGradient.Inputs.handleThe handle to an accumulator.ResourceSparseApplyAdadelta.Inputs.indicesA vector of indices into the first dimension of var and accum.ResourceSparseApplyAdagrad.Inputs.indicesA vector of indices into the first dimension of var and accum.ResourceSparseApplyAdagradDa.Inputs.indicesA vector of indices into the first dimension of var and accum.ResourceSparseApplyAdagradV2.Inputs.indicesA vector of indices into the first dimension of var and accum.ResourceSparseApplyCenteredRmsProp.Inputs.indicesA vector of indices into the first dimension of var, ms and mom.ResourceSparseApplyFtrl.Inputs.indicesA vector of indices into the first dimension of var and accum.ResourceSparseApplyKerasMomentum.Inputs.indicesA vector of indices into the first dimension of var and accum.ResourceSparseApplyMomentum.Inputs.indicesA vector of indices into the first dimension of var and accum.ResourceSparseApplyProximalAdagrad.Inputs.indicesA vector of indices into the first dimension of var and accum.ResourceSparseApplyProximalGradientDescent.Inputs.indicesA vector of indices into the first dimension of var and accum.ResourceSparseApplyRmsProp.Inputs.indicesA vector of indices into the first dimension of var, ms and mom.SparseApplyAdadelta.Inputs.indicesA vector of indices into the first dimension of var and accum.SparseApplyAdagrad.Inputs.indicesA vector of indices into the first dimension of var and accum.SparseApplyAdagradDa.Inputs.indicesA vector of indices into the first dimension of var and accum.SparseApplyCenteredRmsProp.Inputs.indicesA vector of indices into the first dimension of var, ms and mom.SparseApplyFtrl.Inputs.indicesA vector of indices into the first dimension of var and accum.SparseApplyMomentum.Inputs.indicesA vector of indices into the first dimension of var and accum.SparseApplyProximalAdagrad.Inputs.indicesA vector of indices into the first dimension of var and accum.SparseApplyProximalGradientDescent.Inputs.indicesA vector of indices into the first dimension of var and accum.SparseApplyRmsProp.Inputs.indicesA vector of indices into the first dimension of var, ms and mom.PreventGradient.Inputs.inputany tensor.SdcaFprint.Inputs.inputvector of strings to compute fingerprints on.TileGrad.Inputs.inputThe input inputComputeBatchSize.Inputs.inputDatasetThe inputDataset inputApplyAdagradDa.Inputs.l1L1 regularization.ApplyFtrl.Inputs.l1L1 regularization.ApplyProximalAdagrad.Inputs.l1L1 regularization.ApplyProximalGradientDescent.Inputs.l1L1 regularization.ResourceApplyAdagradDa.Inputs.l1L1 regularization.ResourceApplyFtrl.Inputs.l1L1 regularization.ResourceApplyProximalAdagrad.Inputs.l1L1 regularization.ResourceApplyProximalGradientDescent.Inputs.l1L1 regularization.ResourceSparseApplyAdagradDa.Inputs.l1L1 regularization.ResourceSparseApplyFtrl.Inputs.l1L1 regularization.ResourceSparseApplyProximalAdagrad.Inputs.l1L1 regularization.ResourceSparseApplyProximalGradientDescent.Inputs.l1L1 regularization.SparseApplyAdagradDa.Inputs.l1L1 regularization.SparseApplyFtrl.Inputs.l1L1 regularization.SparseApplyProximalAdagrad.Inputs.l1L1 regularization.SparseApplyProximalGradientDescent.Inputs.l1L1 regularization.ApplyAdagradDa.Inputs.l2L2 regularization.ApplyFtrl.Inputs.l2L2 shrinkage regularization.ApplyProximalAdagrad.Inputs.l2L2 regularization.ApplyProximalGradientDescent.Inputs.l2L2 regularization.ResourceApplyAdagradDa.Inputs.l2L2 regularization.ResourceApplyFtrl.Inputs.l2L2 shrinkage regularization.ResourceApplyProximalAdagrad.Inputs.l2L2 regularization.ResourceApplyProximalGradientDescent.Inputs.l2L2 regularization.ResourceSparseApplyAdagradDa.Inputs.l2L2 regularization.ResourceSparseApplyFtrl.Inputs.l2L2 shrinkage regularization.ResourceSparseApplyProximalAdagrad.Inputs.l2L2 regularization.ResourceSparseApplyProximalGradientDescent.Inputs.l2L2 regularization.SparseApplyAdagradDa.Inputs.l2L2 regularization.SparseApplyFtrl.Inputs.l2L2 shrinkage regularization.SparseApplyProximalAdagrad.Inputs.l2L2 regularization.SparseApplyProximalGradientDescent.Inputs.l2L2 regularization.ApplyFtrl.Inputs.l2ShrinkageThe l2Shrinkage inputResourceApplyFtrl.Inputs.l2ShrinkageThe l2Shrinkage inputResourceSparseApplyFtrl.Inputs.l2ShrinkageThe l2Shrinkage inputSparseApplyFtrl.Inputs.l2ShrinkageThe l2Shrinkage inputNegTrain.Inputs.labelsA vector of word ids.ApplyFtrl.Inputs.linearShould be from a Variable().ResourceApplyFtrl.Inputs.linearShould be from a Variable().ResourceSparseApplyFtrl.Inputs.linearShould be from a Variable().SparseApplyFtrl.Inputs.linearShould be from a Variable().AccumulatorApplyGradient.Inputs.localStepThe local_step value at which the gradient was computed.ResourceAccumulatorApplyGradient.Inputs.localStepThe local_step value at which the gradient was computed.ApplyPowerSign.Inputs.logbaseMust be a scalar.ResourceApplyPowerSign.Inputs.logbaseMust be a scalar.ApplyAdadelta.Inputs.lrScaling factor.ApplyAdagrad.Inputs.lrScaling factor.ApplyAdagradDa.Inputs.lrScaling factor.ApplyAdagradV2.Inputs.lrScaling factor.ApplyAdam.Inputs.lrScaling factor.ApplyAdaMax.Inputs.lrScaling factor.ApplyAddSign.Inputs.lrScaling factor.ApplyCenteredRmsProp.Inputs.lrScaling factor.ApplyFtrl.Inputs.lrScaling factor.ApplyMomentum.Inputs.lrScaling factor.ApplyPowerSign.Inputs.lrScaling factor.ApplyProximalAdagrad.Inputs.lrScaling factor.ApplyRmsProp.Inputs.lrScaling factor.NegTrain.Inputs.lrThe lr inputResourceApplyAdadelta.Inputs.lrScaling factor.ResourceApplyAdagrad.Inputs.lrScaling factor.ResourceApplyAdagradDa.Inputs.lrScaling factor.ResourceApplyAdam.Inputs.lrScaling factor.ResourceApplyAdaMax.Inputs.lrScaling factor.ResourceApplyAdamWithAmsgrad.Inputs.lrScaling factor.ResourceApplyAddSign.Inputs.lrScaling factor.ResourceApplyCenteredRmsProp.Inputs.lrScaling factor.ResourceApplyFtrl.Inputs.lrScaling factor.ResourceApplyKerasMomentum.Inputs.lrScaling factor.ResourceApplyMomentum.Inputs.lrScaling factor.ResourceApplyPowerSign.Inputs.lrScaling factor.ResourceApplyProximalAdagrad.Inputs.lrScaling factor.ResourceApplyRmsProp.Inputs.lrScaling factor.ResourceSparseApplyAdadelta.Inputs.lrLearning rate.ResourceSparseApplyAdagrad.Inputs.lrLearning rate.ResourceSparseApplyAdagradDa.Inputs.lrLearning rate.ResourceSparseApplyAdagradV2.Inputs.lrLearning rate.ResourceSparseApplyCenteredRmsProp.Inputs.lrScaling factor.ResourceSparseApplyFtrl.Inputs.lrScaling factor.ResourceSparseApplyKerasMomentum.Inputs.lrLearning rate.ResourceSparseApplyMomentum.Inputs.lrLearning rate.ResourceSparseApplyProximalAdagrad.Inputs.lrLearning rate.ResourceSparseApplyRmsProp.Inputs.lrScaling factor.SparseApplyAdadelta.Inputs.lrLearning rate.SparseApplyAdagrad.Inputs.lrLearning rate.SparseApplyAdagradDa.Inputs.lrLearning rate.SparseApplyCenteredRmsProp.Inputs.lrScaling factor.SparseApplyFtrl.Inputs.lrScaling factor.SparseApplyMomentum.Inputs.lrLearning rate.SparseApplyProximalAdagrad.Inputs.lrLearning rate.SparseApplyRmsProp.Inputs.lrScaling factor.ApplyFtrl.Inputs.lrPowerScaling factor.ResourceApplyFtrl.Inputs.lrPowerScaling factor.ResourceSparseApplyFtrl.Inputs.lrPowerScaling factor.SparseApplyFtrl.Inputs.lrPowerScaling factor.ApplyAdam.Inputs.mShould be from a Variable().ApplyAdaMax.Inputs.mShould be from a Variable().ApplyAddSign.Inputs.mShould be from a Variable().ApplyPowerSign.Inputs.mShould be from a Variable().ResourceApplyAdam.Inputs.mShould be from a Variable().ResourceApplyAdaMax.Inputs.mShould be from a Variable().ResourceApplyAdamWithAmsgrad.Inputs.mShould be from a Variable().ResourceApplyAddSign.Inputs.mShould be from a Variable().ResourceApplyPowerSign.Inputs.mShould be from a Variable().ApplyCenteredRmsProp.Inputs.mgShould be from a Variable().ResourceApplyCenteredRmsProp.Inputs.mgShould be from a Variable().ResourceSparseApplyCenteredRmsProp.Inputs.mgShould be from a Variable().SparseApplyCenteredRmsProp.Inputs.mgShould be from a Variable().ApplyCenteredRmsProp.Inputs.momShould be from a Variable().ApplyRmsProp.Inputs.momShould be from a Variable().ResourceApplyCenteredRmsProp.Inputs.momShould be from a Variable().ResourceApplyRmsProp.Inputs.momShould be from a Variable().ResourceSparseApplyCenteredRmsProp.Inputs.momShould be from a Variable().ResourceSparseApplyRmsProp.Inputs.momShould be from a Variable().SparseApplyCenteredRmsProp.Inputs.momShould be from a Variable().SparseApplyRmsProp.Inputs.momShould be from a Variable().ApplyCenteredRmsProp.Inputs.momentumMomentum Scale.ApplyMomentum.Inputs.momentumMomentum.ApplyRmsProp.Inputs.momentumThe momentum inputResourceApplyCenteredRmsProp.Inputs.momentumMomentum Scale.ResourceApplyKerasMomentum.Inputs.momentumMomentum.ResourceApplyMomentum.Inputs.momentumMomentum.ResourceApplyRmsProp.Inputs.momentumThe momentum inputResourceSparseApplyCenteredRmsProp.Inputs.momentumThe momentum inputResourceSparseApplyKerasMomentum.Inputs.momentumMomentum.ResourceSparseApplyMomentum.Inputs.momentumMomentum.ResourceSparseApplyRmsProp.Inputs.momentumThe momentum inputSparseApplyCenteredRmsProp.Inputs.momentumThe momentum inputSparseApplyMomentum.Inputs.momentumMomentum.SparseApplyRmsProp.Inputs.momentumThe momentum inputApplyCenteredRmsProp.Inputs.msShould be from a Variable().ApplyRmsProp.Inputs.msShould be from a Variable().ResourceApplyCenteredRmsProp.Inputs.msShould be from a Variable().ResourceApplyRmsProp.Inputs.msShould be from a Variable().ResourceSparseApplyCenteredRmsProp.Inputs.msShould be from a Variable().ResourceSparseApplyRmsProp.Inputs.msShould be from a Variable().SparseApplyCenteredRmsProp.Inputs.msShould be from a Variable().SparseApplyRmsProp.Inputs.msShould be from a Variable().TileGrad.Inputs.multiplesThe multiples inputAccumulatorSetGlobalStep.Inputs.newGlobalStepThe new global_step value to set.ResourceAccumulatorSetGlobalStep.Inputs.newGlobalStepThe new global_step value to set.GenerateVocabRemapping.Inputs.newVocabFilePath to the new vocab file.AccumulatorTakeGradient.Inputs.numRequiredNumber of gradients required before we return an aggregate.ResourceAccumulatorTakeGradient.Inputs.numRequiredNumber of gradients required before we return an aggregate.GenerateVocabRemapping.Inputs.oldVocabFilePath to the old vocab file.Restore.Inputs.prefixMust have a single element.Save.Inputs.prefixMust have a single element.ApplyAdadelta.Inputs.rhoDecay factor.ApplyCenteredRmsProp.Inputs.rhoDecay rate.ApplyRmsProp.Inputs.rhoDecay rate.ResourceApplyAdadelta.Inputs.rhoDecay factor.ResourceApplyCenteredRmsProp.Inputs.rhoDecay rate.ResourceApplyRmsProp.Inputs.rhoDecay rate.ResourceSparseApplyAdadelta.Inputs.rhoDecay factor.ResourceSparseApplyCenteredRmsProp.Inputs.rhoDecay rate.ResourceSparseApplyRmsProp.Inputs.rhoDecay rate.SparseApplyAdadelta.Inputs.rhoDecay factor.SparseApplyCenteredRmsProp.Inputs.rhoDecay rate.SparseApplyRmsProp.Inputs.rhoDecay rate.RestoreSlice.Inputs.shapeAndSliceScalar.Restore.Inputs.shapeAndSlicesshape {N}.Save.Inputs.shapeAndSlicesshape {N}.SaveSlices.Inputs.shapesAndSlicesShape[N].ApplyAddSign.Inputs.signDecayMust be a scalar.ApplyPowerSign.Inputs.signDecayMust be a scalar.ResourceApplyAddSign.Inputs.signDecayMust be a scalar.ResourceApplyPowerSign.Inputs.signDecayMust be a scalar.RestoreSlice.Inputs.tensorNameMust have a single element.Restore.Inputs.tensorNamesshape {N}.Save.Inputs.tensorNamesshape {N}.SaveSlices.Inputs.tensorNamesShape[N].ApplyAdam.Inputs.vShould be from a Variable().ApplyAdaMax.Inputs.vShould be from a Variable().ResourceApplyAdam.Inputs.vShould be from a Variable().ResourceApplyAdaMax.Inputs.vShould be from a Variable().ResourceApplyAdamWithAmsgrad.Inputs.vShould be from a Variable().ApplyAdadelta.Inputs.varShould be from a Variable().ApplyAdagrad.Inputs.varShould be from a Variable().ApplyAdagradDa.Inputs.varShould be from a Variable().ApplyAdagradV2.Inputs.varShould be from a Variable().ApplyAdam.Inputs.varShould be from a Variable().ApplyAdaMax.Inputs.varShould be from a Variable().ApplyAddSign.Inputs.varShould be from a Variable().ApplyCenteredRmsProp.Inputs.varShould be from a Variable().ApplyFtrl.Inputs.varShould be from a Variable().ApplyGradientDescent.Inputs.varShould be from a Variable().ApplyMomentum.Inputs.varShould be from a Variable().ApplyPowerSign.Inputs.varShould be from a Variable().ApplyProximalAdagrad.Inputs.varShould be from a Variable().ApplyProximalGradientDescent.Inputs.varShould be from a Variable().ApplyRmsProp.Inputs.varShould be from a Variable().ResourceApplyAdadelta.Inputs.varShould be from a Variable().ResourceApplyAdagrad.Inputs.varShould be from a Variable().ResourceApplyAdagradDa.Inputs.varShould be from a Variable().ResourceApplyAdam.Inputs.varShould be from a Variable().ResourceApplyAdaMax.Inputs.varShould be from a Variable().ResourceApplyAdamWithAmsgrad.Inputs.varShould be from a Variable().ResourceApplyAddSign.Inputs.varShould be from a Variable().ResourceApplyCenteredRmsProp.Inputs.varShould be from a Variable().ResourceApplyFtrl.Inputs.varShould be from a Variable().ResourceApplyGradientDescent.Inputs.varShould be from a Variable().ResourceApplyKerasMomentum.Inputs.varShould be from a Variable().ResourceApplyMomentum.Inputs.varShould be from a Variable().ResourceApplyPowerSign.Inputs.varShould be from a Variable().ResourceApplyProximalAdagrad.Inputs.varShould be from a Variable().ResourceApplyProximalGradientDescent.Inputs.varShould be from a Variable().ResourceApplyRmsProp.Inputs.varShould be from a Variable().ResourceSparseApplyAdadelta.Inputs.varThe var inputResourceSparseApplyAdagrad.Inputs.varShould be from a Variable().ResourceSparseApplyAdagradDa.Inputs.varShould be from a Variable().ResourceSparseApplyAdagradV2.Inputs.varShould be from a Variable().ResourceSparseApplyCenteredRmsProp.Inputs.varShould be from a Variable().ResourceSparseApplyFtrl.Inputs.varShould be from a Variable().ResourceSparseApplyKerasMomentum.Inputs.varShould be from a Variable().ResourceSparseApplyMomentum.Inputs.varShould be from a Variable().ResourceSparseApplyProximalAdagrad.Inputs.varShould be from a Variable().ResourceSparseApplyProximalGradientDescent.Inputs.varShould be from a Variable().ResourceSparseApplyRmsProp.Inputs.varShould be from a Variable().SparseApplyAdadelta.Inputs.varThe var inputSparseApplyAdagrad.Inputs.varShould be from a Variable().SparseApplyAdagradDa.Inputs.varShould be from a Variable().SparseApplyCenteredRmsProp.Inputs.varShould be from a Variable().SparseApplyFtrl.Inputs.varShould be from a Variable().SparseApplyMomentum.Inputs.varShould be from a Variable().SparseApplyProximalAdagrad.Inputs.varShould be from a Variable().SparseApplyProximalGradientDescent.Inputs.varShould be from a Variable().SparseApplyRmsProp.Inputs.varShould be from a Variable().ResourceApplyAdamWithAmsgrad.Inputs.vhatShould be from a Variable().NegTrain.Inputs.wIninput word embedding.NegTrain.Inputs.wOutoutput word embedding.BatchMatMul.Inputs.x2-D or higher with shape[..., r_x, c_x].BatchMatMul.Inputs.y2-D or higher with shape[..., r_y, c_y].Fields in org.tensorflow.op.train with type parameters of type OperandModifier and TypeFieldDescriptionSaveSlices.Inputs.dataNtensors to save.SdcaOptimizer.Inputs.denseFeaturesa list of matrices which contains the dense feature values.SdcaOptimizer.Inputs.denseWeightsa list of vectors where the values are the weights associated with a dense feature group.SymbolicGradient.Inputs.inputa list of input tensors of size N + M;SdcaOptimizer.Inputs.sparseExampleIndicesa list of vectors which contain example indices.SdcaOptimizer.Inputs.sparseFeatureIndicesa list of vectors which contain feature indices.SdcaOptimizer.Inputs.sparseFeatureValuesa list of vectors which contains feature value associated with each feature group.SdcaOptimizer.Inputs.sparseIndicesa list of vectors where each value is the indices which has corresponding weights in sparse_weights.SdcaOptimizer.Inputs.sparseWeightsa list of vectors where each value is the weight associated with a sparse feature group.Save.Inputs.tensorsNtensors to save.SdcaShrinkL1.Inputs.weightsa list of vectors where each value is the weight associated with a feature group.Methods in org.tensorflow.op.train that return types with arguments of type OperandMethods in org.tensorflow.op.train with parameters of type OperandModifier and TypeMethodDescriptionstatic AccumulatorApplyGradientAccumulatorApplyGradient.create(Scope scope, Operand<TString> handle, Operand<TInt64> localStep, Operand<? extends TType> gradient) Factory method to create a class wrapping a new AccumulatorApplyGradient operation.static AccumulatorNumAccumulatedFactory method to create a class wrapping a new AccumulatorNumAccumulated operation.static AccumulatorSetGlobalStepAccumulatorSetGlobalStep.create(Scope scope, Operand<TString> handle, Operand<TInt64> newGlobalStep) Factory method to create a class wrapping a new AccumulatorSetGlobalStep operation.static <T extends TType>
AccumulatorTakeGradient<T> AccumulatorTakeGradient.create(Scope scope, Operand<TString> handle, Operand<TInt32> numRequired, Class<T> dtype) Factory method to create a class wrapping a new AccumulatorTakeGradient operation.static <T extends TType>
ApplyAdadelta<T> ApplyAdadelta.create(Scope scope, Operand<T> var, Operand<T> accum, Operand<T> accumUpdate, Operand<T> lr, Operand<T> rho, Operand<T> epsilon, Operand<T> grad, ApplyAdadelta.Options... options) Factory method to create a class wrapping a new ApplyAdadelta operation.static <T extends TType>
ApplyAdagrad<T> ApplyAdagrad.create(Scope scope, Operand<T> var, Operand<T> accum, Operand<T> lr, Operand<T> grad, ApplyAdagrad.Options... options) Factory method to create a class wrapping a new ApplyAdagrad operation.static <T extends TType>
ApplyAdagradDa<T> ApplyAdagradDa.create(Scope scope, Operand<T> var, Operand<T> gradientAccumulator, Operand<T> gradientSquaredAccumulator, Operand<T> grad, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<TInt64> globalStep, ApplyAdagradDa.Options... options) Factory method to create a class wrapping a new ApplyAdagradDA operation.static <T extends TType>
ApplyAdagradV2<T> ApplyAdagradV2.create(Scope scope, Operand<T> var, Operand<T> accum, Operand<T> lr, Operand<T> epsilon, Operand<T> grad, ApplyAdagradV2.Options... options) Factory method to create a class wrapping a new ApplyAdagradV2 operation.ApplyAdam.create(Scope scope, Operand<T> var, Operand<T> m, Operand<T> v, Operand<T> beta1Power, Operand<T> beta2Power, Operand<T> lr, Operand<T> beta1, Operand<T> beta2, Operand<T> epsilon, Operand<T> grad, ApplyAdam.Options... options) Factory method to create a class wrapping a new ApplyAdam operation.static <T extends TType>
ApplyAdaMax<T> ApplyAdaMax.create(Scope scope, Operand<T> var, Operand<T> m, Operand<T> v, Operand<T> beta1Power, Operand<T> lr, Operand<T> beta1, Operand<T> beta2, Operand<T> epsilon, Operand<T> grad, ApplyAdaMax.Options... options) Factory method to create a class wrapping a new ApplyAdaMax operation.static <T extends TType>
ApplyAddSign<T> ApplyAddSign.create(Scope scope, Operand<T> var, Operand<T> m, Operand<T> lr, Operand<T> alpha, Operand<T> signDecay, Operand<T> beta, Operand<T> grad, ApplyAddSign.Options... options) Factory method to create a class wrapping a new ApplyAddSign operation.static <T extends TType>
ApplyCenteredRmsProp<T> ApplyCenteredRmsProp.create(Scope scope, Operand<T> var, Operand<T> mg, Operand<T> ms, Operand<T> mom, Operand<T> lr, Operand<T> rho, Operand<T> momentum, Operand<T> epsilon, Operand<T> grad, ApplyCenteredRmsProp.Options... options) Factory method to create a class wrapping a new ApplyCenteredRMSProp operation.ApplyFtrl.create(Scope scope, Operand<T> var, Operand<T> accum, Operand<T> linear, Operand<T> grad, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<T> l2Shrinkage, Operand<T> lrPower, ApplyFtrl.Options... options) Factory method to create a class wrapping a new ApplyFtrlV2 operation.static <T extends TType>
ApplyGradientDescent<T> ApplyGradientDescent.create(Scope scope, Operand<T> var, Operand<T> alpha, Operand<T> delta, ApplyGradientDescent.Options... options) Factory method to create a class wrapping a new ApplyGradientDescent operation.static <T extends TType>
ApplyMomentum<T> ApplyMomentum.create(Scope scope, Operand<T> var, Operand<T> accum, Operand<T> lr, Operand<T> grad, Operand<T> momentum, ApplyMomentum.Options... options) Factory method to create a class wrapping a new ApplyMomentum operation.static <T extends TType>
ApplyPowerSign<T> ApplyPowerSign.create(Scope scope, Operand<T> var, Operand<T> m, Operand<T> lr, Operand<T> logbase, Operand<T> signDecay, Operand<T> beta, Operand<T> grad, ApplyPowerSign.Options... options) Factory method to create a class wrapping a new ApplyPowerSign operation.static <T extends TType>
ApplyProximalAdagrad<T> ApplyProximalAdagrad.create(Scope scope, Operand<T> var, Operand<T> accum, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<T> grad, ApplyProximalAdagrad.Options... options) Factory method to create a class wrapping a new ApplyProximalAdagrad operation.static <T extends TType>
ApplyProximalGradientDescent<T> ApplyProximalGradientDescent.create(Scope scope, Operand<T> var, Operand<T> alpha, Operand<T> l1, Operand<T> l2, Operand<T> delta, ApplyProximalGradientDescent.Options... options) Factory method to create a class wrapping a new ApplyProximalGradientDescent operation.static <T extends TType>
ApplyRmsProp<T> ApplyRmsProp.create(Scope scope, Operand<T> var, Operand<T> ms, Operand<T> mom, Operand<T> lr, Operand<T> rho, Operand<T> momentum, Operand<T> epsilon, Operand<T> grad, ApplyRmsProp.Options... options) Factory method to create a class wrapping a new ApplyRMSProp operation.static <V extends TType>
BatchMatMul<V> BatchMatMul.create(Scope scope, Operand<? extends TType> x, Operand<? extends TType> y, Class<V> Tout, BatchMatMul.Options... options) Factory method to create a class wrapping a new BatchMatMulV3 operation.static ComputeBatchSizeFactory method to create a class wrapping a new ComputeBatchSize operation.static DistributedSaveDistributedSave.create(Scope scope, Operand<? extends TType> dataset, Operand<TString> directory, Operand<TString> address, DistributedSave.Options... options) Factory method to create a class wrapping a new DistributedSave operation.static GenerateVocabRemappingGenerateVocabRemapping.create(Scope scope, Operand<TString> newVocabFile, Operand<TString> oldVocabFile, Long newVocabOffset, Long numNewVocab, GenerateVocabRemapping.Options... options) Factory method to create a class wrapping a new GenerateVocabRemapping operation.static MergeV2CheckpointsMergeV2Checkpoints.create(Scope scope, Operand<TString> checkpointPrefixes, Operand<TString> destinationPrefix, MergeV2Checkpoints.Options... options) Factory method to create a class wrapping a new MergeV2Checkpoints operation.static NegTrainNegTrain.create(Scope scope, Operand<TFloat32> wIn, Operand<TFloat32> wOut, Operand<TInt32> examples, Operand<TInt32> labels, Operand<TFloat32> lr, List<Long> vocabCount, Long numNegativeSamples) Factory method to create a class wrapping a new NegTrain operation.static <T extends TType>
PreventGradient<T> PreventGradient.create(Scope scope, Operand<T> input, PreventGradient.Options... options) Factory method to create a class wrapping a new PreventGradient operation.ResourceAccumulatorApplyGradient.create(Scope scope, Operand<? extends TType> handle, Operand<TInt64> localStep, Operand<? extends TType> gradient) Factory method to create a class wrapping a new ResourceAccumulatorApplyGradient operation.Factory method to create a class wrapping a new ResourceAccumulatorNumAccumulated operation.ResourceAccumulatorSetGlobalStep.create(Scope scope, Operand<? extends TType> handle, Operand<TInt64> newGlobalStep) Factory method to create a class wrapping a new ResourceAccumulatorSetGlobalStep operation.static <T extends TType>
ResourceAccumulatorTakeGradient<T> ResourceAccumulatorTakeGradient.create(Scope scope, Operand<? extends TType> handle, Operand<TInt32> numRequired, Class<T> dtype) Factory method to create a class wrapping a new ResourceAccumulatorTakeGradient operation.static <T extends TType>
ResourceApplyAdadeltaResourceApplyAdadelta.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> accum, Operand<? extends TType> accumUpdate, Operand<T> lr, Operand<T> rho, Operand<T> epsilon, Operand<T> grad, ResourceApplyAdadelta.Options... options) Factory method to create a class wrapping a new ResourceApplyAdadelta operation.static <T extends TType>
ResourceApplyAdagradResourceApplyAdagrad.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> accum, Operand<T> lr, Operand<T> epsilon, Operand<T> grad, ResourceApplyAdagrad.Options... options) Factory method to create a class wrapping a new ResourceApplyAdagradV2 operation.static <T extends TType>
ResourceApplyAdagradDaResourceApplyAdagradDa.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> gradientAccumulator, Operand<? extends TType> gradientSquaredAccumulator, Operand<T> grad, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<TInt64> globalStep, ResourceApplyAdagradDa.Options... options) Factory method to create a class wrapping a new ResourceApplyAdagradDA operation.static <T extends TType>
ResourceApplyAdamResourceApplyAdam.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> m, Operand<? extends TType> v, Operand<T> beta1Power, Operand<T> beta2Power, Operand<T> lr, Operand<T> beta1, Operand<T> beta2, Operand<T> epsilon, Operand<T> grad, ResourceApplyAdam.Options... options) Factory method to create a class wrapping a new ResourceApplyAdam operation.static <T extends TType>
ResourceApplyAdaMaxResourceApplyAdaMax.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> m, Operand<? extends TType> v, Operand<T> beta1Power, Operand<T> lr, Operand<T> beta1, Operand<T> beta2, Operand<T> epsilon, Operand<T> grad, ResourceApplyAdaMax.Options... options) Factory method to create a class wrapping a new ResourceApplyAdaMax operation.static <T extends TType>
ResourceApplyAdamWithAmsgradResourceApplyAdamWithAmsgrad.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> m, Operand<? extends TType> v, Operand<? extends TType> vhat, Operand<T> beta1Power, Operand<T> beta2Power, Operand<T> lr, Operand<T> beta1, Operand<T> beta2, Operand<T> epsilon, Operand<T> grad, ResourceApplyAdamWithAmsgrad.Options... options) Factory method to create a class wrapping a new ResourceApplyAdamWithAmsgrad operation.static <T extends TType>
ResourceApplyAddSignResourceApplyAddSign.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> m, Operand<T> lr, Operand<T> alpha, Operand<T> signDecay, Operand<T> beta, Operand<T> grad, ResourceApplyAddSign.Options... options) Factory method to create a class wrapping a new ResourceApplyAddSign operation.static <T extends TType>
ResourceApplyCenteredRmsPropResourceApplyCenteredRmsProp.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> mg, Operand<? extends TType> ms, Operand<? extends TType> mom, Operand<T> lr, Operand<T> rho, Operand<T> momentum, Operand<T> epsilon, Operand<T> grad, ResourceApplyCenteredRmsProp.Options... options) Factory method to create a class wrapping a new ResourceApplyCenteredRMSProp operation.static <T extends TType>
ResourceApplyFtrlResourceApplyFtrl.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> accum, Operand<? extends TType> linear, Operand<T> grad, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<T> l2Shrinkage, Operand<T> lrPower, ResourceApplyFtrl.Options... options) Factory method to create a class wrapping a new ResourceApplyFtrlV2 operation.static <T extends TType>
ResourceApplyGradientDescentResourceApplyGradientDescent.create(Scope scope, Operand<? extends TType> var, Operand<T> alpha, Operand<T> delta, ResourceApplyGradientDescent.Options... options) Factory method to create a class wrapping a new ResourceApplyGradientDescent operation.static <T extends TType>
ResourceApplyKerasMomentumResourceApplyKerasMomentum.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> accum, Operand<T> lr, Operand<T> grad, Operand<T> momentum, ResourceApplyKerasMomentum.Options... options) Factory method to create a class wrapping a new ResourceApplyKerasMomentum operation.static <T extends TType>
ResourceApplyMomentumResourceApplyMomentum.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> accum, Operand<T> lr, Operand<T> grad, Operand<T> momentum, ResourceApplyMomentum.Options... options) Factory method to create a class wrapping a new ResourceApplyMomentum operation.static <T extends TType>
ResourceApplyPowerSignResourceApplyPowerSign.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> m, Operand<T> lr, Operand<T> logbase, Operand<T> signDecay, Operand<T> beta, Operand<T> grad, ResourceApplyPowerSign.Options... options) Factory method to create a class wrapping a new ResourceApplyPowerSign operation.static <T extends TType>
ResourceApplyProximalAdagradResourceApplyProximalAdagrad.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> accum, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<T> grad, ResourceApplyProximalAdagrad.Options... options) Factory method to create a class wrapping a new ResourceApplyProximalAdagrad operation.static <T extends TType>
ResourceApplyProximalGradientDescentResourceApplyProximalGradientDescent.create(Scope scope, Operand<? extends TType> var, Operand<T> alpha, Operand<T> l1, Operand<T> l2, Operand<T> delta, ResourceApplyProximalGradientDescent.Options... options) Factory method to create a class wrapping a new ResourceApplyProximalGradientDescent operation.static <T extends TType>
ResourceApplyRmsPropResourceApplyRmsProp.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> ms, Operand<? extends TType> mom, Operand<T> lr, Operand<T> rho, Operand<T> momentum, Operand<T> epsilon, Operand<T> grad, ResourceApplyRmsProp.Options... options) Factory method to create a class wrapping a new ResourceApplyRMSProp operation.static <T extends TType>
ResourceSparseApplyAdadeltaResourceSparseApplyAdadelta.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> accum, Operand<? extends TType> accumUpdate, Operand<T> lr, Operand<T> rho, Operand<T> epsilon, Operand<T> grad, Operand<? extends TNumber> indices, ResourceSparseApplyAdadelta.Options... options) Factory method to create a class wrapping a new ResourceSparseApplyAdadelta operation.static <T extends TType>
ResourceSparseApplyAdagradResourceSparseApplyAdagrad.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> accum, Operand<T> lr, Operand<T> grad, Operand<? extends TNumber> indices, ResourceSparseApplyAdagrad.Options... options) Factory method to create a class wrapping a new ResourceSparseApplyAdagrad operation.static <T extends TType>
ResourceSparseApplyAdagradDaResourceSparseApplyAdagradDa.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> gradientAccumulator, Operand<? extends TType> gradientSquaredAccumulator, Operand<T> grad, Operand<? extends TNumber> indices, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<TInt64> globalStep, ResourceSparseApplyAdagradDa.Options... options) Factory method to create a class wrapping a new ResourceSparseApplyAdagradDA operation.static <T extends TType>
ResourceSparseApplyAdagradV2ResourceSparseApplyAdagradV2.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> accum, Operand<T> lr, Operand<T> epsilon, Operand<T> grad, Operand<? extends TNumber> indices, ResourceSparseApplyAdagradV2.Options... options) Factory method to create a class wrapping a new ResourceSparseApplyAdagradV2 operation.static <T extends TType>
ResourceSparseApplyCenteredRmsPropResourceSparseApplyCenteredRmsProp.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> mg, Operand<? extends TType> ms, Operand<? extends TType> mom, Operand<T> lr, Operand<T> rho, Operand<T> momentum, Operand<T> epsilon, Operand<T> grad, Operand<? extends TNumber> indices, ResourceSparseApplyCenteredRmsProp.Options... options) Factory method to create a class wrapping a new ResourceSparseApplyCenteredRMSProp operation.static <T extends TType>
ResourceSparseApplyFtrlResourceSparseApplyFtrl.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> accum, Operand<? extends TType> linear, Operand<T> grad, Operand<? extends TNumber> indices, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<T> l2Shrinkage, Operand<T> lrPower, ResourceSparseApplyFtrl.Options... options) Factory method to create a class wrapping a new ResourceSparseApplyFtrlV2 operation.static <T extends TType>
ResourceSparseApplyKerasMomentumResourceSparseApplyKerasMomentum.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> accum, Operand<T> lr, Operand<T> grad, Operand<? extends TNumber> indices, Operand<T> momentum, ResourceSparseApplyKerasMomentum.Options... options) Factory method to create a class wrapping a new ResourceSparseApplyKerasMomentum operation.static <T extends TType>
ResourceSparseApplyMomentumResourceSparseApplyMomentum.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> accum, Operand<T> lr, Operand<T> grad, Operand<? extends TNumber> indices, Operand<T> momentum, ResourceSparseApplyMomentum.Options... options) Factory method to create a class wrapping a new ResourceSparseApplyMomentum operation.static <T extends TType>
ResourceSparseApplyProximalAdagradResourceSparseApplyProximalAdagrad.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> accum, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<T> grad, Operand<? extends TNumber> indices, ResourceSparseApplyProximalAdagrad.Options... options) Factory method to create a class wrapping a new ResourceSparseApplyProximalAdagrad operation.static <T extends TType>
ResourceSparseApplyProximalGradientDescentResourceSparseApplyProximalGradientDescent.create(Scope scope, Operand<? extends TType> var, Operand<T> alpha, Operand<T> l1, Operand<T> l2, Operand<T> grad, Operand<? extends TNumber> indices, ResourceSparseApplyProximalGradientDescent.Options... options) Factory method to create a class wrapping a new ResourceSparseApplyProximalGradientDescent operation.static <T extends TType>
ResourceSparseApplyRmsPropResourceSparseApplyRmsProp.create(Scope scope, Operand<? extends TType> var, Operand<? extends TType> ms, Operand<? extends TType> mom, Operand<T> lr, Operand<T> rho, Operand<T> momentum, Operand<T> epsilon, Operand<T> grad, Operand<? extends TNumber> indices, ResourceSparseApplyRmsProp.Options... options) Factory method to create a class wrapping a new ResourceSparseApplyRMSProp operation.static RestoreRestore.create(Scope scope, Operand<TString> prefix, Operand<TString> tensorNames, Operand<TString> shapeAndSlices, List<Class<? extends TType>> dtypes) Factory method to create a class wrapping a new RestoreV2 operation.static <T extends TType>
RestoreSlice<T> RestoreSlice.create(Scope scope, Operand<TString> filePattern, Operand<TString> tensorName, Operand<TString> shapeAndSlice, Class<T> dt, RestoreSlice.Options... options) Factory method to create a class wrapping a new RestoreSlice operation.static SaveSave.create(Scope scope, Operand<TString> prefix, Operand<TString> tensorNames, Operand<TString> shapeAndSlices, Iterable<Operand<?>> tensors) Factory method to create a class wrapping a new SaveV2 operation.static SaveSlicesSaveSlices.create(Scope scope, Operand<TString> filename, Operand<TString> tensorNames, Operand<TString> shapesAndSlices, Iterable<Operand<?>> data) Factory method to create a class wrapping a new SaveSlices operation.static SdcaFprintFactory method to create a class wrapping a new SdcaFprint operation.static SdcaOptimizerSdcaOptimizer.create(Scope scope, Iterable<Operand<TInt64>> sparseExampleIndices, Iterable<Operand<TInt64>> sparseFeatureIndices, Iterable<Operand<TFloat32>> sparseFeatureValues, Iterable<Operand<TFloat32>> denseFeatures, Operand<TFloat32> exampleWeights, Operand<TFloat32> exampleLabels, Iterable<Operand<TInt64>> sparseIndices, Iterable<Operand<TFloat32>> sparseWeights, Iterable<Operand<TFloat32>> denseWeights, Operand<TFloat32> exampleStateData, String lossType, Float l1, Float l2, Long numLossPartitions, Long numInnerIterations, SdcaOptimizer.Options... options) Factory method to create a class wrapping a new SdcaOptimizerV2 operation.static <T extends TType>
SparseApplyAdadelta<T> SparseApplyAdadelta.create(Scope scope, Operand<T> var, Operand<T> accum, Operand<T> accumUpdate, Operand<T> lr, Operand<T> rho, Operand<T> epsilon, Operand<T> grad, Operand<? extends TNumber> indices, SparseApplyAdadelta.Options... options) Factory method to create a class wrapping a new SparseApplyAdadelta operation.static <T extends TType>
SparseApplyAdagrad<T> SparseApplyAdagrad.create(Scope scope, Operand<T> var, Operand<T> accum, Operand<T> lr, Operand<T> epsilon, Operand<T> grad, Operand<? extends TNumber> indices, SparseApplyAdagrad.Options... options) Factory method to create a class wrapping a new SparseApplyAdagradV2 operation.static <T extends TType>
SparseApplyAdagradDa<T> SparseApplyAdagradDa.create(Scope scope, Operand<T> var, Operand<T> gradientAccumulator, Operand<T> gradientSquaredAccumulator, Operand<T> grad, Operand<? extends TNumber> indices, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<TInt64> globalStep, SparseApplyAdagradDa.Options... options) Factory method to create a class wrapping a new SparseApplyAdagradDA operation.static <T extends TType>
SparseApplyCenteredRmsProp<T> SparseApplyCenteredRmsProp.create(Scope scope, Operand<T> var, Operand<T> mg, Operand<T> ms, Operand<T> mom, Operand<T> lr, Operand<T> rho, Operand<T> momentum, Operand<T> epsilon, Operand<T> grad, Operand<? extends TNumber> indices, SparseApplyCenteredRmsProp.Options... options) Factory method to create a class wrapping a new SparseApplyCenteredRMSProp operation.static <T extends TType>
SparseApplyFtrl<T> SparseApplyFtrl.create(Scope scope, Operand<T> var, Operand<T> accum, Operand<T> linear, Operand<T> grad, Operand<? extends TNumber> indices, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<T> l2Shrinkage, Operand<T> lrPower, SparseApplyFtrl.Options... options) Factory method to create a class wrapping a new SparseApplyFtrlV2 operation.static <T extends TType>
SparseApplyMomentum<T> SparseApplyMomentum.create(Scope scope, Operand<T> var, Operand<T> accum, Operand<T> lr, Operand<T> grad, Operand<? extends TNumber> indices, Operand<T> momentum, SparseApplyMomentum.Options... options) Factory method to create a class wrapping a new SparseApplyMomentum operation.static <T extends TType>
SparseApplyProximalAdagrad<T> SparseApplyProximalAdagrad.create(Scope scope, Operand<T> var, Operand<T> accum, Operand<T> lr, Operand<T> l1, Operand<T> l2, Operand<T> grad, Operand<? extends TNumber> indices, SparseApplyProximalAdagrad.Options... options) Factory method to create a class wrapping a new SparseApplyProximalAdagrad operation.static <T extends TType>
SparseApplyProximalGradientDescent<T> SparseApplyProximalGradientDescent.create(Scope scope, Operand<T> var, Operand<T> alpha, Operand<T> l1, Operand<T> l2, Operand<T> grad, Operand<? extends TNumber> indices, SparseApplyProximalGradientDescent.Options... options) Factory method to create a class wrapping a new SparseApplyProximalGradientDescent operation.static <T extends TType>
SparseApplyRmsProp<T> SparseApplyRmsProp.create(Scope scope, Operand<T> var, Operand<T> ms, Operand<T> mom, Operand<T> lr, Operand<T> rho, Operand<T> momentum, Operand<T> epsilon, Operand<T> grad, Operand<? extends TNumber> indices, SparseApplyRmsProp.Options... options) Factory method to create a class wrapping a new SparseApplyRMSProp operation.Factory method to create a class wrapping a new TileGrad operation.Method parameters in org.tensorflow.op.train with type arguments of type OperandModifier and TypeMethodDescriptionstatic SaveSave.create(Scope scope, Operand<TString> prefix, Operand<TString> tensorNames, Operand<TString> shapeAndSlices, Iterable<Operand<?>> tensors) Factory method to create a class wrapping a new SaveV2 operation.static SaveSlicesSaveSlices.create(Scope scope, Operand<TString> filename, Operand<TString> tensorNames, Operand<TString> shapesAndSlices, Iterable<Operand<?>> data) Factory method to create a class wrapping a new SaveSlices operation.static SdcaOptimizerSdcaOptimizer.create(Scope scope, Iterable<Operand<TInt64>> sparseExampleIndices, Iterable<Operand<TInt64>> sparseFeatureIndices, Iterable<Operand<TFloat32>> sparseFeatureValues, Iterable<Operand<TFloat32>> denseFeatures, Operand<TFloat32> exampleWeights, Operand<TFloat32> exampleLabels, Iterable<Operand<TInt64>> sparseIndices, Iterable<Operand<TFloat32>> sparseWeights, Iterable<Operand<TFloat32>> denseWeights, Operand<TFloat32> exampleStateData, String lossType, Float l1, Float l2, Long numLossPartitions, Long numInnerIterations, SdcaOptimizer.Options... options) Factory method to create a class wrapping a new SdcaOptimizerV2 operation.static SdcaShrinkL1Factory method to create a class wrapping a new SdcaShrinkL1 operation.static SymbolicGradientSymbolicGradient.create(Scope scope, Iterable<Operand<?>> input, List<Class<? extends TType>> Tout, ConcreteFunction f) Factory method to create a class wrapping a new SymbolicGradient operation. -
Uses of Operand in org.tensorflow.op.xla
Classes in org.tensorflow.op.xla that implement OperandModifier and TypeClassDescriptionfinal classConcats input tensor across all dimensions.final classXlaRecvFromHost<T extends TType>An op to receive a tensor from the host. output: the tensor that will be received from the host.final classReceives deduplication data (indices and weights) from the embedding core.final classThe XlaSparseCoreSgd operationfinal classThe XlaSparseDenseMatmulGradWithSgdAndCsrInput operationfinal classThe XlaSparseDenseMatmulGradWithSgdAndStaticBufferSize operationfinal classThe XlaSparseDenseMatmulWithCsrInput operationfinal classThe XlaSparseDenseMatmulWithStaticBufferSize operationClasses in org.tensorflow.op.xla that implement interfaces with type arguments of type OperandModifier and TypeClassDescriptionfinal classReadVariableSplitND<T extends TType>Splits resource variable input tensor across all dimensions.final classSplits input tensor across all dimensions.final classA pseudo-op to represent host-side computation in an XLA program.final classAn op that receives embedding activations on the TPU.final classThe XlaSparseDenseMatmulGradWithCsrInput operationFields in org.tensorflow.op.xla declared as OperandModifier and TypeFieldDescriptionXlaSparseCoreAdagrad.Inputs.accumulatorThe accumulator inputXlaSparseCoreAdagradMomentum.Inputs.accumulatorThe accumulator inputXlaSparseCoreFtrl.Inputs.accumulatorThe accumulator inputXlaSparseDenseMatmulGradWithAdagradAndCsrInput.Inputs.accumulatorThe accumulator inputXlaSparseDenseMatmulGradWithAdagradAndStaticBufferSize.Inputs.accumulatorThe accumulator inputXlaSparseDenseMatmulGradWithAdagradMomentumAndCsrInput.Inputs.accumulatorThe accumulator inputXlaSparseDenseMatmulGradWithAdagradMomentumAndStaticBufferSize.Inputs.accumulatorThe accumulator inputXlaSparseDenseMatmulGradWithFtrlAndCsrInput.Inputs.accumulatorThe accumulator inputXlaSparseDenseMatmulGradWithFtrlAndStaticBufferSize.Inputs.accumulatorThe accumulator inputXlaSparseDenseMatmulGradWithAdagradAndCsrInput.Inputs.activationGradientsThe activationGradients inputXlaSparseDenseMatmulGradWithAdagradAndStaticBufferSize.Inputs.activationGradientsThe activationGradients inputXlaSparseDenseMatmulGradWithAdagradMomentumAndCsrInput.Inputs.activationGradientsThe activationGradients inputXlaSparseDenseMatmulGradWithAdagradMomentumAndStaticBufferSize.Inputs.activationGradientsThe activationGradients inputXlaSparseDenseMatmulGradWithAdamAndCsrInput.Inputs.activationGradientsThe activationGradients inputXlaSparseDenseMatmulGradWithAdamAndStaticBufferSize.Inputs.activationGradientsThe activationGradients inputXlaSparseDenseMatmulGradWithCsrInput.Inputs.activationGradientsThe activationGradients inputXlaSparseDenseMatmulGradWithFtrlAndCsrInput.Inputs.activationGradientsThe activationGradients inputXlaSparseDenseMatmulGradWithFtrlAndStaticBufferSize.Inputs.activationGradientsThe activationGradients inputXlaSparseDenseMatmulGradWithSgdAndCsrInput.Inputs.activationGradientsThe activationGradients inputXlaSparseDenseMatmulGradWithSgdAndStaticBufferSize.Inputs.activationGradientsThe activationGradients inputXlaSparseCoreFtrl.Inputs.betaThe beta inputXlaSparseCoreAdagradMomentum.Inputs.beta1The beta1 inputXlaSparseCoreAdam.Inputs.beta1The beta1 inputXlaSparseCoreAdam.Inputs.beta2The beta2 inputXlaSparseDenseMatmul.Inputs.colIdsThe colIds inputXlaRecvTPUEmbeddingActivations.Inputs.deduplicationDataA Tensor with type=DT_VARIANT containing the deduplication data.XlaSendTPUEmbeddingGradients.Inputs.deduplicationDataA Tensor with type=DT_VARIANT containing the deduplication data.XlaSparseCoreAdagrad.Inputs.embeddingTableThe embeddingTable inputXlaSparseCoreAdagradMomentum.Inputs.embeddingTableThe embeddingTable inputXlaSparseCoreAdam.Inputs.embeddingTableThe embeddingTable inputXlaSparseCoreFtrl.Inputs.embeddingTableThe embeddingTable inputXlaSparseCoreSgd.Inputs.embeddingTableThe embeddingTable inputXlaSparseDenseMatmul.Inputs.embeddingTableThe embeddingTable inputXlaSparseDenseMatmulGradWithAdagradAndCsrInput.Inputs.embeddingTableThe embeddingTable inputXlaSparseDenseMatmulGradWithAdagradAndStaticBufferSize.Inputs.embeddingTableThe embeddingTable inputXlaSparseDenseMatmulGradWithAdagradMomentumAndCsrInput.Inputs.embeddingTableThe embeddingTable inputXlaSparseDenseMatmulGradWithAdagradMomentumAndStaticBufferSize.Inputs.embeddingTableThe embeddingTable inputXlaSparseDenseMatmulGradWithAdamAndCsrInput.Inputs.embeddingTableThe embeddingTable inputXlaSparseDenseMatmulGradWithAdamAndStaticBufferSize.Inputs.embeddingTableThe embeddingTable inputXlaSparseDenseMatmulGradWithFtrlAndCsrInput.Inputs.embeddingTableThe embeddingTable inputXlaSparseDenseMatmulGradWithFtrlAndStaticBufferSize.Inputs.embeddingTableThe embeddingTable inputXlaSparseDenseMatmulGradWithSgdAndCsrInput.Inputs.embeddingTableThe embeddingTable inputXlaSparseDenseMatmulGradWithSgdAndStaticBufferSize.Inputs.embeddingTableThe embeddingTable inputXlaSparseDenseMatmulWithCsrInput.Inputs.embeddingTableThe embeddingTable inputXlaSparseDenseMatmulWithStaticBufferSize.Inputs.embeddingTableThe embeddingTable inputXlaSparseCoreAdagradMomentum.Inputs.epsilonThe epsilon inputXlaSparseCoreAdam.Inputs.epsilonThe epsilon inputXlaSparseCoreAdagrad.Inputs.gradientThe gradient inputXlaSparseCoreAdagradMomentum.Inputs.gradientThe gradient inputXlaSparseCoreAdam.Inputs.gradientThe gradient inputXlaSparseCoreFtrl.Inputs.gradientThe gradient inputXlaSparseCoreSgd.Inputs.gradientThe gradient inputXlaSparseCoreAdagrad.Inputs.indicesThe indices inputXlaSparseCoreAdagradMomentum.Inputs.indicesThe indices inputXlaSparseCoreAdam.Inputs.indicesThe indices inputXlaSparseCoreFtrl.Inputs.indicesThe indices inputXlaSparseCoreSgd.Inputs.indicesThe indices inputSplitND.Inputs.inputInput tensor to split across all dimensions.XlaSendToHost.Inputs.inputThe input inputXlaSparseCoreFtrl.Inputs.l2RegularizationStrengthThe l2RegularizationStrength inputXlaSparseCoreAdagrad.Inputs.learningRateThe learningRate inputXlaSparseCoreAdagradMomentum.Inputs.learningRateThe learningRate inputXlaSparseCoreAdam.Inputs.learningRateThe learningRate inputXlaSparseCoreFtrl.Inputs.learningRateThe learningRate inputXlaSparseCoreSgd.Inputs.learningRateThe learningRate inputXlaSparseDenseMatmulGradWithAdagradAndCsrInput.Inputs.learningRateThe learningRate inputXlaSparseDenseMatmulGradWithAdagradAndStaticBufferSize.Inputs.learningRateThe learningRate inputXlaSparseDenseMatmulGradWithAdagradMomentumAndCsrInput.Inputs.learningRateThe learningRate inputXlaSparseDenseMatmulGradWithAdagradMomentumAndStaticBufferSize.Inputs.learningRateThe learningRate inputXlaSparseDenseMatmulGradWithAdamAndCsrInput.Inputs.learningRateThe learningRate inputXlaSparseDenseMatmulGradWithAdamAndStaticBufferSize.Inputs.learningRateThe learningRate inputXlaSparseDenseMatmulGradWithFtrlAndCsrInput.Inputs.learningRateThe learningRate inputXlaSparseDenseMatmulGradWithFtrlAndStaticBufferSize.Inputs.learningRateThe learningRate inputXlaSparseDenseMatmulGradWithSgdAndCsrInput.Inputs.learningRateThe learningRate inputXlaSparseDenseMatmulGradWithSgdAndStaticBufferSize.Inputs.learningRateThe learningRate inputXlaSparseCoreFtrl.Inputs.learningRatePowerThe learningRatePower inputXlaSparseCoreFtrl.Inputs.linearThe linear inputXlaSparseDenseMatmulGradWithFtrlAndCsrInput.Inputs.linearThe linear inputXlaSparseDenseMatmulGradWithFtrlAndStaticBufferSize.Inputs.linearThe linear inputXlaSparseDenseMatmulGradWithAdagradMomentumAndCsrInput.Inputs.momentaThe momenta inputXlaSparseDenseMatmulGradWithAdagradMomentumAndStaticBufferSize.Inputs.momentaThe momenta inputXlaSparseDenseMatmulGradWithAdamAndCsrInput.Inputs.momentaThe momenta inputXlaSparseDenseMatmulGradWithAdamAndStaticBufferSize.Inputs.momentaThe momenta inputXlaSparseCoreAdagradMomentum.Inputs.momentumThe momentum inputXlaSparseCoreAdam.Inputs.momentumThe momentum inputXlaSparseDenseMatmulGradWithAdagradAndCsrInput.Inputs.numMinibatchesPerPhysicalSparseCoreThe numMinibatchesPerPhysicalSparseCore inputXlaSparseDenseMatmulGradWithAdagradAndStaticBufferSize.Inputs.numMinibatchesPerPhysicalSparseCoreThe numMinibatchesPerPhysicalSparseCore inputXlaSparseDenseMatmulGradWithAdagradMomentumAndCsrInput.Inputs.numMinibatchesPerPhysicalSparseCoreThe numMinibatchesPerPhysicalSparseCore inputXlaSparseDenseMatmulGradWithAdagradMomentumAndStaticBufferSize.Inputs.numMinibatchesPerPhysicalSparseCoreThe numMinibatchesPerPhysicalSparseCore inputXlaSparseDenseMatmulGradWithAdamAndCsrInput.Inputs.numMinibatchesPerPhysicalSparseCoreThe numMinibatchesPerPhysicalSparseCore inputXlaSparseDenseMatmulGradWithAdamAndStaticBufferSize.Inputs.numMinibatchesPerPhysicalSparseCoreThe numMinibatchesPerPhysicalSparseCore inputXlaSparseDenseMatmulGradWithCsrInput.Inputs.numMinibatchesPerPhysicalSparseCoreThe numMinibatchesPerPhysicalSparseCore inputXlaSparseDenseMatmulGradWithFtrlAndCsrInput.Inputs.numMinibatchesPerPhysicalSparseCoreThe numMinibatchesPerPhysicalSparseCore inputXlaSparseDenseMatmulGradWithFtrlAndStaticBufferSize.Inputs.numMinibatchesPerPhysicalSparseCoreThe numMinibatchesPerPhysicalSparseCore inputXlaSparseDenseMatmulGradWithSgdAndCsrInput.Inputs.numMinibatchesPerPhysicalSparseCoreThe numMinibatchesPerPhysicalSparseCore inputXlaSparseDenseMatmulGradWithSgdAndStaticBufferSize.Inputs.numMinibatchesPerPhysicalSparseCoreThe numMinibatchesPerPhysicalSparseCore inputXlaSparseDenseMatmulWithCsrInput.Inputs.numMinibatchesPerPhysicalSparseCoreThe numMinibatchesPerPhysicalSparseCore inputXlaSparseDenseMatmulWithStaticBufferSize.Inputs.numMinibatchesPerPhysicalSparseCoreThe numMinibatchesPerPhysicalSparseCore inputXlaSparseDenseMatmul.Inputs.offsetsThe offsets inputAssignVariableConcatND.Inputs.resourceResource variable for concatenated input tensors across all dimensions.ReadVariableSplitND.Inputs.resourceResource variable of input tensor to split across all dimensions.XlaSparseDenseMatmul.Inputs.rowIdsThe rowIds inputXlaSparseDenseMatmulGradWithAdagradAndCsrInput.Inputs.rowPointersThe rowPointers inputXlaSparseDenseMatmulGradWithAdagradAndStaticBufferSize.Inputs.rowPointersThe rowPointers inputXlaSparseDenseMatmulGradWithAdagradMomentumAndCsrInput.Inputs.rowPointersThe rowPointers inputXlaSparseDenseMatmulGradWithAdagradMomentumAndStaticBufferSize.Inputs.rowPointersThe rowPointers inputXlaSparseDenseMatmulGradWithAdamAndCsrInput.Inputs.rowPointersThe rowPointers inputXlaSparseDenseMatmulGradWithAdamAndStaticBufferSize.Inputs.rowPointersThe rowPointers inputXlaSparseDenseMatmulGradWithCsrInput.Inputs.rowPointersThe rowPointers inputXlaSparseDenseMatmulGradWithFtrlAndCsrInput.Inputs.rowPointersThe rowPointers inputXlaSparseDenseMatmulGradWithFtrlAndStaticBufferSize.Inputs.rowPointersThe rowPointers inputXlaSparseDenseMatmulGradWithSgdAndCsrInput.Inputs.rowPointersThe rowPointers inputXlaSparseDenseMatmulGradWithSgdAndStaticBufferSize.Inputs.rowPointersThe rowPointers inputXlaSparseDenseMatmulWithCsrInput.Inputs.rowPointersThe rowPointers inputXlaSparseDenseMatmulWithStaticBufferSize.Inputs.rowPointersThe rowPointers inputXlaSparseDenseMatmulGradWithAdagradAndCsrInput.Inputs.sortedGainsThe sortedGains inputXlaSparseDenseMatmulGradWithAdagradAndStaticBufferSize.Inputs.sortedGainsThe sortedGains inputXlaSparseDenseMatmulGradWithAdagradMomentumAndCsrInput.Inputs.sortedGainsThe sortedGains inputXlaSparseDenseMatmulGradWithAdagradMomentumAndStaticBufferSize.Inputs.sortedGainsThe sortedGains inputXlaSparseDenseMatmulGradWithAdamAndCsrInput.Inputs.sortedGainsThe sortedGains inputXlaSparseDenseMatmulGradWithAdamAndStaticBufferSize.Inputs.sortedGainsThe sortedGains inputXlaSparseDenseMatmulGradWithCsrInput.Inputs.sortedGainsThe sortedGains inputXlaSparseDenseMatmulGradWithFtrlAndCsrInput.Inputs.sortedGainsThe sortedGains inputXlaSparseDenseMatmulGradWithFtrlAndStaticBufferSize.Inputs.sortedGainsThe sortedGains inputXlaSparseDenseMatmulGradWithSgdAndCsrInput.Inputs.sortedGainsThe sortedGains inputXlaSparseDenseMatmulGradWithSgdAndStaticBufferSize.Inputs.sortedGainsThe sortedGains inputXlaSparseDenseMatmulWithCsrInput.Inputs.sortedGainsThe sortedGains inputXlaSparseDenseMatmulWithStaticBufferSize.Inputs.sortedGainsThe sortedGains inputXlaSparseDenseMatmulGradWithAdagradAndCsrInput.Inputs.sortedSampleIdsThe sortedSampleIds inputXlaSparseDenseMatmulGradWithAdagradAndStaticBufferSize.Inputs.sortedSampleIdsThe sortedSampleIds inputXlaSparseDenseMatmulGradWithAdagradMomentumAndCsrInput.Inputs.sortedSampleIdsThe sortedSampleIds inputXlaSparseDenseMatmulGradWithAdagradMomentumAndStaticBufferSize.Inputs.sortedSampleIdsThe sortedSampleIds inputXlaSparseDenseMatmulGradWithAdamAndCsrInput.Inputs.sortedSampleIdsThe sortedSampleIds inputXlaSparseDenseMatmulGradWithAdamAndStaticBufferSize.Inputs.sortedSampleIdsThe sortedSampleIds inputXlaSparseDenseMatmulGradWithCsrInput.Inputs.sortedSampleIdsThe sortedSampleIds inputXlaSparseDenseMatmulGradWithFtrlAndCsrInput.Inputs.sortedSampleIdsThe sortedSampleIds inputXlaSparseDenseMatmulGradWithFtrlAndStaticBufferSize.Inputs.sortedSampleIdsThe sortedSampleIds inputXlaSparseDenseMatmulGradWithSgdAndCsrInput.Inputs.sortedSampleIdsThe sortedSampleIds inputXlaSparseDenseMatmulGradWithSgdAndStaticBufferSize.Inputs.sortedSampleIdsThe sortedSampleIds inputXlaSparseDenseMatmulWithCsrInput.Inputs.sortedSampleIdsThe sortedSampleIds inputXlaSparseDenseMatmulWithStaticBufferSize.Inputs.sortedSampleIdsThe sortedSampleIds inputXlaSparseDenseMatmulGradWithAdagradAndCsrInput.Inputs.sortedTokenIdsThe sortedTokenIds inputXlaSparseDenseMatmulGradWithAdagradAndStaticBufferSize.Inputs.sortedTokenIdsThe sortedTokenIds inputXlaSparseDenseMatmulGradWithAdagradMomentumAndCsrInput.Inputs.sortedTokenIdsThe sortedTokenIds inputXlaSparseDenseMatmulGradWithAdagradMomentumAndStaticBufferSize.Inputs.sortedTokenIdsThe sortedTokenIds inputXlaSparseDenseMatmulGradWithAdamAndCsrInput.Inputs.sortedTokenIdsThe sortedTokenIds inputXlaSparseDenseMatmulGradWithAdamAndStaticBufferSize.Inputs.sortedTokenIdsThe sortedTokenIds inputXlaSparseDenseMatmulGradWithCsrInput.Inputs.sortedTokenIdsThe sortedTokenIds inputXlaSparseDenseMatmulGradWithFtrlAndCsrInput.Inputs.sortedTokenIdsThe sortedTokenIds inputXlaSparseDenseMatmulGradWithFtrlAndStaticBufferSize.Inputs.sortedTokenIdsThe sortedTokenIds inputXlaSparseDenseMatmulGradWithSgdAndCsrInput.Inputs.sortedTokenIdsThe sortedTokenIds inputXlaSparseDenseMatmulGradWithSgdAndStaticBufferSize.Inputs.sortedTokenIdsThe sortedTokenIds inputXlaSparseDenseMatmulWithCsrInput.Inputs.sortedTokenIdsThe sortedTokenIds inputXlaSparseDenseMatmulWithStaticBufferSize.Inputs.sortedTokenIdsThe sortedTokenIds inputXlaSparseDenseMatmul.Inputs.valuesThe values inputXlaSparseCoreAdam.Inputs.velocityThe velocity inputXlaSparseDenseMatmulGradWithAdamAndCsrInput.Inputs.velocityThe velocity inputXlaSparseDenseMatmulGradWithAdamAndStaticBufferSize.Inputs.velocityThe velocity inputFields in org.tensorflow.op.xla with type parameters of type OperandModifier and TypeFieldDescriptionXlaSendTPUEmbeddingGradients.Inputs.gradientsA TensorList of gradients with which to update embedding tables.XlaSparseDenseMatmulGradWithCsrInput.Inputs.hyperparametersThe hyperparameters inputAssignVariableConcatND.Inputs.inputsInput tensor slices in row-major order to merge across all dimensions.ConcatND.Inputs.inputsInput tensor slices in row-major order to merge across all dimensions.XlaHostCompute.Inputs.inputsA list of tensors that will be sent to the host.XlaSendTPUEmbeddingGradients.Inputs.learningRatesA TensorList of learning rates used for updating the embedding tables via the optimizer.XlaSparseDenseMatmulGradWithCsrInput.Inputs.tablesThe tables inputMethods in org.tensorflow.op.xla that return types with arguments of type OperandModifier and TypeMethodDescriptionReadVariableSplitND.iterator()SplitND.iterator()XlaHostCompute.iterator()XlaRecvTPUEmbeddingActivations.iterator()XlaSparseDenseMatmulGradWithCsrInput.iterator()Methods in org.tensorflow.op.xla with parameters of type OperandModifier and TypeMethodDescriptionstatic AssignVariableConcatNDAssignVariableConcatND.create(Scope scope, Operand<? extends TType> resource, Iterable<Operand<? extends TType>> inputs, List<Long> numConcats, AssignVariableConcatND.Options... options) Factory method to create a class wrapping a new AssignVariableXlaConcatND operation.static <T extends TType>
ReadVariableSplitND<T> ReadVariableSplitND.create(Scope scope, Operand<? extends TType> resource, Class<T> T, Long N, List<Long> numSplits, ReadVariableSplitND.Options... options) Factory method to create a class wrapping a new ReadVariableXlaSplitND operation.SplitND.create(Scope scope, Operand<T> input, Long N, List<Long> numSplits, SplitND.Options... options) Factory method to create a class wrapping a new XlaSplitND operation.XlaRecvTPUEmbeddingActivations.create(Scope scope, Operand<? extends TType> deduplicationData, Long numTables, String config, String embeddingPartitions, String hbmBuffersConfig, String tpuTopology) Factory method to create a class wrapping a new XlaRecvTPUEmbeddingActivationsV2 operation.static XlaSendToHostFactory method to create a class wrapping a new XlaSendToHost operation.static XlaSendTPUEmbeddingGradientsXlaSendTPUEmbeddingGradients.create(Scope scope, Iterable<Operand<TFloat32>> gradients, Iterable<Operand<TFloat32>> learningRates, Operand<? extends TType> deduplicationData, String config, String embeddingPartitions, String hbmBuffersConfig, String tpuTopology, XlaSendTPUEmbeddingGradients.Options... options) Factory method to create a class wrapping a new XlaSendTPUEmbeddingGradientsV2 operation.static XlaSparseCoreAdagradXlaSparseCoreAdagrad.create(Scope scope, Operand<TInt32> indices, Operand<TFloat32> gradient, Operand<TFloat32> learningRate, Operand<TFloat32> accumulator, Operand<TFloat32> embeddingTable, Long featureWidth) Factory method to create a class wrapping a new XlaSparseCoreAdagrad operation.static XlaSparseCoreAdagradMomentumXlaSparseCoreAdagradMomentum.create(Scope scope, Operand<TInt32> indices, Operand<TFloat32> gradient, Operand<TFloat32> learningRate, Operand<TFloat32> beta1, Operand<TFloat32> epsilon, Operand<TFloat32> accumulator, Operand<TFloat32> momentum, Operand<TFloat32> embeddingTable, Long featureWidth, Boolean useNesterov, Float beta2, Float exponent) Factory method to create a class wrapping a new XlaSparseCoreAdagradMomentum operation.static XlaSparseCoreAdamXlaSparseCoreAdam.create(Scope scope, Operand<TFloat32> embeddingTable, Operand<TInt32> indices, Operand<TFloat32> gradient, Operand<TFloat32> learningRate, Operand<TFloat32> momentum, Operand<TFloat32> velocity, Operand<TFloat32> beta1, Operand<TFloat32> beta2, Operand<TFloat32> epsilon, Long featureWidth, Boolean useSumInsideSqrt) Factory method to create a class wrapping a new XlaSparseCoreAdam operation.static XlaSparseCoreFtrlXlaSparseCoreFtrl.create(Scope scope, Operand<TFloat32> embeddingTable, Operand<TFloat32> accumulator, Operand<TFloat32> linear, Operand<TFloat32> learningRate, Operand<TInt32> indices, Operand<TFloat32> gradient, Operand<TFloat32> beta, Operand<TFloat32> learningRatePower, Operand<TFloat32> l2RegularizationStrength, Long featureWidth, Boolean multiplyLinearByLearningRate, Float l1RegularizationStrength) Factory method to create a class wrapping a new XlaSparseCoreFtrl operation.static XlaSparseCoreSgdXlaSparseCoreSgd.create(Scope scope, Operand<TInt32> indices, Operand<TFloat32> gradient, Operand<TFloat32> learningRate, Operand<TFloat32> embeddingTable, Long featureWidth) Factory method to create a class wrapping a new XlaSparseCoreSgd operation.static XlaSparseDenseMatmulXlaSparseDenseMatmul.create(Scope scope, Operand<TInt32> rowIds, Operand<? extends TType> colIds, Operand<TFloat32> values, Operand<? extends TType> offsets, Operand<TFloat32> embeddingTable, Long maxIdsPerPartition, Long maxUniqueIdsPerPartition, Long inputSize) Factory method to create a class wrapping a new XlaSparseDenseMatmul operation.XlaSparseDenseMatmulGradWithAdagradAndCsrInput.create(Scope scope, Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> activationGradients, Operand<TFloat32> learningRate, Operand<TFloat32> embeddingTable, Operand<TFloat32> accumulator, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, String tableName, XlaSparseDenseMatmulGradWithAdagradAndCsrInput.Options... options) Factory method to create a class wrapping a new XlaSparseDenseMatmulGradWithAdagradAndCsrInput operation.XlaSparseDenseMatmulGradWithAdagradAndStaticBufferSize.create(Scope scope, Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> activationGradients, Operand<TFloat32> learningRate, Operand<TFloat32> embeddingTable, Operand<TFloat32> accumulator, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, Long maxIdsPerSparseCore, Long maxUniqueIdsPerSparseCore, String tableName, XlaSparseDenseMatmulGradWithAdagradAndStaticBufferSize.Options... options) Factory method to create a class wrapping a new XlaSparseDenseMatmulGradWithAdagradAndStaticBufferSize operation.XlaSparseDenseMatmulGradWithAdagradMomentumAndCsrInput.create(Scope scope, Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> activationGradients, Operand<TFloat32> learningRate, Operand<TFloat32> embeddingTable, Operand<TFloat32> accumulator, Operand<TFloat32> momenta, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, Boolean useNesterov, Float exponent, Float beta1, Float beta2, Float epsilon, String tableName, XlaSparseDenseMatmulGradWithAdagradMomentumAndCsrInput.Options... options) Factory method to create a class wrapping a new XlaSparseDenseMatmulGradWithAdagradMomentumAndCsrInput operation.XlaSparseDenseMatmulGradWithAdagradMomentumAndStaticBufferSize.create(Scope scope, Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> activationGradients, Operand<TFloat32> learningRate, Operand<TFloat32> embeddingTable, Operand<TFloat32> accumulator, Operand<TFloat32> momenta, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, Boolean useNesterov, Float exponent, Float beta1, Float beta2, Float epsilon, Long maxIdsPerSparseCore, Long maxUniqueIdsPerSparseCore, String tableName, XlaSparseDenseMatmulGradWithAdagradMomentumAndStaticBufferSize.Options... options) Factory method to create a class wrapping a new XlaSparseDenseMatmulGradWithAdagradMomentumAndStaticBufferSize operation.XlaSparseDenseMatmulGradWithAdamAndCsrInput.create(Scope scope, Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> activationGradients, Operand<TFloat32> learningRate, Operand<TFloat32> embeddingTable, Operand<TFloat32> momenta, Operand<TFloat32> velocity, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, Boolean useSumInsideSqrt, Float beta1, Float beta2, Float epsilon, String tableName, XlaSparseDenseMatmulGradWithAdamAndCsrInput.Options... options) Factory method to create a class wrapping a new XlaSparseDenseMatmulGradWithAdamAndCsrInput operation.XlaSparseDenseMatmulGradWithAdamAndStaticBufferSize.create(Scope scope, Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> activationGradients, Operand<TFloat32> learningRate, Operand<TFloat32> embeddingTable, Operand<TFloat32> momenta, Operand<TFloat32> velocity, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, Boolean useSumInsideSqrt, Float beta1, Float beta2, Float epsilon, Long maxIdsPerSparseCore, Long maxUniqueIdsPerSparseCore, String tableName, XlaSparseDenseMatmulGradWithAdamAndStaticBufferSize.Options... options) Factory method to create a class wrapping a new XlaSparseDenseMatmulGradWithAdamAndStaticBufferSize operation.XlaSparseDenseMatmulGradWithCsrInput.create(Scope scope, Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> activationGradients, Iterable<Operand<TFloat32>> tables, Iterable<Operand<TFloat32>> hyperparameters, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, ConcreteFunction customComputation, String tableName) Factory method to create a class wrapping a new XlaSparseDenseMatmulGradWithCsrInput operation.XlaSparseDenseMatmulGradWithFtrlAndCsrInput.create(Scope scope, Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> activationGradients, Operand<TFloat32> learningRate, Operand<TFloat32> embeddingTable, Operand<TFloat32> accumulator, Operand<TFloat32> linear, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, Boolean multiplyLinearByLearningRate, Float beta, Float learningRatePower, Float l1RegularizationStrength, Float l2RegularizationStrength, String tableName, XlaSparseDenseMatmulGradWithFtrlAndCsrInput.Options... options) Factory method to create a class wrapping a new XlaSparseDenseMatmulGradWithFtrlAndCsrInput operation.XlaSparseDenseMatmulGradWithFtrlAndStaticBufferSize.create(Scope scope, Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> activationGradients, Operand<TFloat32> learningRate, Operand<TFloat32> embeddingTable, Operand<TFloat32> accumulator, Operand<TFloat32> linear, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, Boolean multiplyLinearByLearningRate, Float beta, Float learningRatePower, Float l1RegularizationStrength, Float l2RegularizationStrength, Long maxIdsPerSparseCore, Long maxUniqueIdsPerSparseCore, String tableName, XlaSparseDenseMatmulGradWithFtrlAndStaticBufferSize.Options... options) Factory method to create a class wrapping a new XlaSparseDenseMatmulGradWithFtrlAndStaticBufferSize operation.XlaSparseDenseMatmulGradWithSgdAndCsrInput.create(Scope scope, Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> activationGradients, Operand<TFloat32> learningRate, Operand<TFloat32> embeddingTable, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, String tableName, XlaSparseDenseMatmulGradWithSgdAndCsrInput.Options... options) Factory method to create a class wrapping a new XlaSparseDenseMatmulGradWithSgdAndCsrInput operation.XlaSparseDenseMatmulGradWithSgdAndStaticBufferSize.create(Scope scope, Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> activationGradients, Operand<TFloat32> learningRate, Operand<TFloat32> embeddingTable, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, Long maxIdsPerSparseCore, Long maxUniqueIdsPerSparseCore, String tableName, XlaSparseDenseMatmulGradWithSgdAndStaticBufferSize.Options... options) Factory method to create a class wrapping a new XlaSparseDenseMatmulGradWithSgdAndStaticBufferSize operation.XlaSparseDenseMatmulWithCsrInput.create(Scope scope, Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> embeddingTable, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, Long inputSize, Float quantizationConfigLow, Float quantizationConfigHigh, Long quantizationConfigNumBuckets, String tableName) Factory method to create a class wrapping a new XlaSparseDenseMatmulWithCsrInput operation.XlaSparseDenseMatmulWithStaticBufferSize.create(Scope scope, Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> embeddingTable, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, Long inputSize, Float quantizationConfigLow, Float quantizationConfigHigh, Long quantizationConfigNumBuckets, Long maxIdsPerSparseCore, Long maxUniqueIdsPerSparseCore, String tableName) Factory method to create a class wrapping a new XlaSparseDenseMatmulWithStaticBufferSize operation.Method parameters in org.tensorflow.op.xla with type arguments of type OperandModifier and TypeMethodDescriptionstatic AssignVariableConcatNDAssignVariableConcatND.create(Scope scope, Operand<? extends TType> resource, Iterable<Operand<? extends TType>> inputs, List<Long> numConcats, AssignVariableConcatND.Options... options) Factory method to create a class wrapping a new AssignVariableXlaConcatND operation.ConcatND.create(Scope scope, Iterable<Operand<T>> inputs, List<Long> numConcats, ConcatND.Options... options) Factory method to create a class wrapping a new XlaConcatND operation.static XlaHostComputeXlaHostCompute.create(Scope scope, Iterable<Operand<?>> inputs, List<Class<? extends TType>> Toutputs, List<String> ancestors, List<Shape> shapes, ConcreteFunction shapeInferenceGraph, String key, XlaHostCompute.Options... options) Factory method to create a class wrapping a new XlaHostCompute operation.static XlaSendTPUEmbeddingGradientsXlaSendTPUEmbeddingGradients.create(Scope scope, Iterable<Operand<TFloat32>> gradients, Iterable<Operand<TFloat32>> learningRates, Operand<? extends TType> deduplicationData, String config, String embeddingPartitions, String hbmBuffersConfig, String tpuTopology, XlaSendTPUEmbeddingGradients.Options... options) Factory method to create a class wrapping a new XlaSendTPUEmbeddingGradientsV2 operation.XlaSparseDenseMatmulGradWithCsrInput.create(Scope scope, Operand<TInt32> rowPointers, Operand<TInt32> sortedSampleIds, Operand<TInt32> sortedTokenIds, Operand<TFloat32> sortedGains, Operand<TFloat32> activationGradients, Iterable<Operand<TFloat32>> tables, Iterable<Operand<TFloat32>> hyperparameters, Operand<TInt32> numMinibatchesPerPhysicalSparseCore, ConcreteFunction customComputation, String tableName) Factory method to create a class wrapping a new XlaSparseDenseMatmulGradWithCsrInput operation.
NcclAllReduceinstead